{ "NousResearch/hermes-agent": { "summary": "Hermes Agent is an AI agent with a learning loop that creates and improves skills, stores and searches knowledge across sessions, and builds a model of the user over time. It supports several model providers, terminal interfaces, messaging platforms, scheduled automations, subagents, and multiple terminal backends.", "highlights": [ "Creates skills from experience and improves them during use.", "Supports Telegram, Discord, Slack, WhatsApp, Signal, and CLI.", "Includes local, Docker, SSH, Modal, Daytona, and Vercel Sandbox backends." ], "readmeHash": "38b023e1bb33bf302dbc7d0856dd7d79e87cbd67de506dac47a3fed6e5c776bc", "generatedAt": "2026-10-22T21:48:03.558Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "NousResearch/Hermes-Function-Calling": { "summary": "Hermes-Function-Calling provides a framework for the Hermes Pro Large Language Model to execute specific tasks through structured tool use. It utilizes a recursive loop to generate function calls, execute them via Python scripts, and process the returned data. The repository supports both standard function calling and JSON mode inference using Pydantic models for schema validation. Users can extend the system by integrating custom functions and tools, such as financial data retrieval via the yfinance library.", "highlights": [ "Supports function calling and JSON mode inference", "Allows custom tool integration via Python scripts", "Uses Pydantic for structured JSON schema validation" ], "readmeHash": "0bcf440317d1fe4fff8fc4c9c69aa6972591adb80dbc48486b0f456d1ad569da", "generatedAt": "2026-06-15T10:11:06.530Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:00:21.383Z" }, "NousResearch/hermes-agent-self-evolution": { "summary": "Hermes Agent Self-Evolution is a framework for the automatic improvement of Hermes Agent's capabilities. It utilizes DSPy and Genetic-Pareto Prompt Evolution (GEPA) to optimize skills, prompts, and code through reflective evolutionary search. The system analyzes execution traces to propose targeted mutations without requiring GPU training. All evolved variants are subject to strict guardrails, including test suites and size limits, before undergoing human review via pull requests.", "highlights": [ "Optimizes skills and prompts using DSPy and GEPA", "No GPU training required; operates via API calls", "Strict guardrails including full test suites and size limits" ], "readmeHash": "ebecc0a1f483e655041b9aa531ef7535ee0165f7058b65420896d6e89c7acc92", "generatedAt": "2026-06-23T07:46:45.233Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:00:33.245Z" }, "NousResearch/hermes-paperclip-adapter": { "summary": "The hermes-paperclip-adapter is a bridge that integrates Hermes Agent into the Paperclip company management system as a managed employee. It functions by wrapping the Hermes CLI, parsing raw terminal output into structured transcript entries and clean markdown for the Paperclip UI. This allows the agent to respond to task assignments and issue comments while utilizing its native suite of 30+ tools and persistent memory. The adapter also manages session state, skill synchronization, and multi-provider inference settings directly within the Paperclip environment.", "highlights": [ "Integrates Hermes Agent as a managed employee in Paperclip systems", "Parses raw CLI output into structured tool cards and markdown", "Supports 8 inference providers including Anthropic, OpenAI, and OpenRouter" ], "readmeHash": "4c9894eb6c4262b531cc9589d53b6b04abbbb317f420a5dd2e2f09316d1172ce", "generatedAt": "2026-06-23T07:46:50.192Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:00:42.054Z" }, "NousResearch/autonovel": { "summary": "Autonovel is an autonomous pipeline designed to generate complete, print-ready novels from a seed concept. It employs a modify-evaluate-keep/discard loop across four phases, moving from world-building and drafting to automated revision and final export. The system integrates a dual-persona review process using Claude Opus to refine prose and structure. The pipeline can produce a variety of final formats, including PDFs, ePubs, audiobooks, and landing pages.", "highlights": [ "Multi-phase pipeline from seed concept to print-ready PDF/ePub", "Dual-persona review loop using Claude Opus for literary critique", "Integrated art generation and ElevenLabs multi-voice audiobook production" ], "readmeHash": "6d7dc5597f0d2d5e5a6a444176369a28ce9415764eea430cfa38e86667380719", "generatedAt": "2026-06-15T10:11:36.162Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:00:44.999Z" }, "outsourc-e/hermes-workspace": { "summary": "Hermes Workspace is a comprehensive command center designed to orchestrate AI agents through a unified web interface. It integrates with the vanilla hermes-agent to provide a zero-fork environment for managing chat, memory, and system operations. The platform features a Swarm Mode that utilizes tmux-backed workers for autonomous, role-based task dispatching and a Kanban board for tracking progress. Users can manage over 2,000 skills, browse MCP catalogs, and access a cross-platform PTY terminal from a single dashboard.", "highlights": [ "Swarm Mode with role-based dispatch and tmux-backed workers", "Integrated memory browser, skills manager, and MCP catalog", "Full workspace file browser with Monaco and PTY terminal" ], "readmeHash": "72114091d80b86f5ddef8f78428c055cbcf913ee5b823c8f52db596755b85e82", "generatedAt": "2026-06-06T07:31:45.413Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:00:47.594Z" }, "fathah/hermes-desktop": { "summary": "Hermes One is a community-maintained native desktop app for installing, configuring, and chatting with Hermes Agent. It uses the official Hermes install script, stores Hermes in `~/.hermes`, and provides a GUI for chat, sessions, profiles, memory, skills, tools, scheduling, and messaging gateways.", "highlights": [ "Uses the official Hermes install script.", "Provides a GUI for managing profiles, sessions, skills, and tools.", "Available from the project website or GitHub Releases." ], "readmeHash": "4451e84b75e0cb22fff9e74763235aed3ff6afc95c493bbd894f6f08907fef5c", "generatedAt": "2026-10-12T17:06:45.261Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:02.858Z" }, "sanchomuzax/hermes-webui": { "summary": "Hermes WebUI is a graphical dashboard designed to monitor and manage the Hermes Agent ecosystem. It utilizes a FastAPI backend and a React frontend to interface with the agent's existing data files via a WebSocket-based polling bridge. This approach allows for real-time visibility into sessions, costs, and scheduled jobs without requiring modifications to the core agent code. Users can inspect conversation histories, manage skills, and view system configurations through a responsive interface.", "highlights": [ "Real-time dashboard for session activity and cost tracking", "Full inspection of conversation histories and skill source code", "Read-only monitoring of configuration and environment variables" ], "readmeHash": "083245c60a81c8cd3f8bcdeedf7369691efa4389f86788d3da8854582f9e06c9", "generatedAt": "2026-07-14T20:09:20.540Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:27:29.943Z" }, "Euraika-Labs/pan-ui": { "summary": "Pan is a self-hosted web dashboard and workspace designed specifically for the Hermes Agent runtime. It provides a centralized interface to manage agent interactions through a chat-first environment that supports SSE streaming and tool timelines. Users can discover and install skills from skills.sh, manage MCP integrations, and organize work into isolated profiles with unique memory and sessions. The system includes a setup wizard that automatically detects Hermes configurations and can run as a background daemon or system service.", "highlights": [ "Chat-first workspace with real-time SSE streaming and tool approval cards", "Integrated marketplace for discovering and installing skills and MCP servers", "Isolated profile management for separate sessions, memory, and API keys" ], "readmeHash": "f0c28f4a223760a75a2c524dd27a3fd72b9db1abaa8c591893dcb71f717cac52", "generatedAt": "2026-05-11T09:08:46.681Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:00:58.018Z" }, "mukul975/Anthropic-Cybersecurity-Skills": { "summary": "Anthropic-Cybersecurity-Skills is an open-source library containing 818 structured cybersecurity skills designed for AI agents. The repository maps these skills across six industry frameworks, including MITRE ATT&CK, NIST CSF 2.0, and MITRE ATLAS, following the agentskills.io open standard.", "highlights": [ "Contains 818 structured skills across 34 security domains.", "Maps skills to MITRE ATT&CK, ATLAS, D3FEND, and F3 frameworks.", "Compatible with over 26 AI platforms." ], "readmeHash": "1593e862b0997724c33225af83e45bf5138f01130ac7d1c3e2338e5aa12ab4e6", "generatedAt": "2026-08-31T07:58:48.408Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-31T15:33:25.003Z" }, "conorbronsdon/avoid-ai-writing": { "summary": "avoid-ai-writing is a portable writing skill for compatible AI agents that audits and rewrites English prose for AI-writing patterns. It offers rewrite, detect-only, and in-place editing modes, with optional voice profiles and a bundled detector.", "highlights": [ "Rewrite mode supports up to two editing passes.", "In-place editing refuses source code, configuration, and generated data.", "Detector can score files or piped text." ], "readmeHash": "225a3dad131c13d83592f6263b98dcb17fedb651f0b50e04b092b18dd8525be2", "generatedAt": "2026-10-15T21:34:38.361Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:02.536Z" }, "wondelai/skills": { "summary": "Wondelai Skills is a library of 65 agent skills and metaskills compatible with agentskills.io-compliant agents, including Claude Code, Codex, and Cursor. The library provides expert frameworks based on industry style guides and books, alongside guided metaskills that maintain state in a project's docs folder to orchestrate tasks across sessions.", "highlights": [ "Includes 51 expert frameworks and 14 guided metaskills.", "Compatible with Claude Code, Codex, and agent-plugins.org specifications.", "Maintains session state within the project's docs folder." ], "readmeHash": "9f4d810c69195b92b23bdc3b03cde9f5f7e0fe77c5a3cb533d6bca7179f08535", "generatedAt": "2026-09-14T14:31:03.388Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:36:00.242Z" }, "DougTrajano/pydantic-ai-skills": { "summary": "pydantic-ai-skills is a companion package for pydantic-ai-harness that enables remote skill registries and bundled-file execution for Pydantic AI agents. It implements progressive disclosure, allowing agents to load full instructions, reference documents, and scripts only when a specific task requires them.", "highlights": [ "Supports remote skill registries via Git and S3 sources.", "Provides tools for reading bundled resources and executing scripts.", "Includes sandboxed execution to isolate untrusted scripts from the host." ], "readmeHash": "d6bf0f92ebada774c092931111ac40685154c73a8b9864003dcfa1c5d37d86f5", "generatedAt": "2026-09-06T04:51:07.357Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-07T13:43:15.132Z" }, "Agents365-ai/drawio-skill": { "summary": "drawio-skill is an Agent Skills-format skill that turns natural language descriptions and sources such as Terraform, Kubernetes, SQL DDL, and codebases into editable .drawio diagrams. It supports incremental synchronization that preserves manual layout, architecture rules enforced in CI, and export to formats including HTML, PowerPoint, Mermaid, and Markdown. It runs through a CLI called diagramctl, with an optional MCP server for MCP hosts.", "highlights": [ "Generates editable .drawio files from prompts and real sources", "Incremental sync preserves manual layout and annotations", "CLI diagramctl with optional MCP server" ], "readmeHash": "c13324a7f46f3f568a0472fde97efd62951864eb6abff9ab05cd14eb27717736", "generatedAt": "2026-10-12T00:26:02.406Z", "model": "deepseek/deepseek-v4.1-flash", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:03:45.690Z" }, "smartcontractkit/chainlink-agent-skills": { "summary": "This repository collects official Chainlink coding skills, each following the Agent Skills specification. It lists nine skills covering areas such as CRE, CCIP, Data Feeds, Data Streams, ACE, VRF, and a confidential AI attester. Skills can be installed with the vercel skills CLI, either at project level or globally with the -g flag.", "highlights": [ "Nine Chainlink skills listed, each following the Agent Skills spec", "Install via npx skills add smartcontractkit/chainlink-agent-skills", "README states repo contains community examples only" ], "readmeHash": "3ff7c41dc3da2fbe3545fa447125edd187f8877840562c87b8e9ff99aaea647e", "generatedAt": "2026-10-12T00:26:02.733Z", "model": "deepseek/deepseek-v4.1-flash", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:03:46.723Z" }, "tlehman/litprog-skill": { "summary": "The litprog-skill is a Claude Code extension that implements Donald Knuth's literate programming paradigm by transforming codebases into human-readable narratives. It works by generating a .lit.md file that weaves together prose, Mermaid diagrams, and LaTeX math with the original source code. The tool includes a tangler to extract runnable files and a reverse-sync engine to ensure the documentation stays updated when source files are edited. This approach helps developers document architectural decisions and facilitates easier onboarding through structured, essay-like code explanations.", "highlights": [ "Generates .lit.md files combining prose, diagrams, and source code", "Extracts runnable source files from documentation using a tangler", "Automatically syncs source file edits back to the literate document" ], "readmeHash": "f8dace63feb38870d4f849f1bd78a3798115ba153b8fdb16c2807671cafee6bd", "generatedAt": "2026-04-17T18:15:41.478Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:01:22.567Z" }, "esaradev/icarus-plugin": { "summary": "Icarus is a Hermes plugin that provides agents with shared memory and a pipeline for creating replacement models. It captures agent decisions and work into a markdown-based directory, allowing different agent profiles to access a common knowledge base. This captured data is then filtered by quality and exported as training pairs for fine-tuning cheaper models via Together AI. Users can evaluate these candidate models and switch to them once they meet specific performance thresholds.", "highlights": [ "Cross-instance shared memory via markdown files", "Automated capture of high-value decisions and outcomes", "Pipeline for fine-tuning and switching to replacement models" ], "readmeHash": "8b13e6b60d76b2d252da08e1118227fd7dda81aaca0025a3d2f64f6008ab82b4", "generatedAt": "2026-06-08T09:40:27.899Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:01:25.464Z" }, "chigwell/skilldock.io": { "summary": "SkillDock is an OpenAPI-driven Python SDK and CLI for interacting with the SkillDock API. It enables users to search, install, and manage skill releases, and supports commerce flows via TON.", "highlights": [ "Provides a Python SkilldockClient and a CLI", "Supports skill installation with recursive dependency resolution", "Includes commerce features using TON payments" ], "readmeHash": "9beb150914e814b27f173931a0761e3e33e9328472c64771d5d0c309a2fdf730", "generatedAt": "2026-07-14T21:30:43.016Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:37:01.993Z" }, "black-forest-labs/skills": { "summary": "BFL Skills provides tools to teach coding agents how to prompt FLUX models and integrate with the BFL API following the agentskills.io specification. The project includes specialized modules for image prompting best practices and a suite of video generation skills.", "highlights": [ "Supports FLUX image and video generation prompting", "Includes BFL API integration for polling and webhooks", "Provides a FLUX 3 video suite with eight specialist skills" ], "readmeHash": "66340f6d577920e2edd182fac61fde7936cda8ea92d0b7bb2c79ff0fcd9d8d9a", "generatedAt": "2026-09-14T14:31:04.676Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:35:52.787Z" }, "Romanescu11/hermes-skill-factory": { "summary": "Skill Factory is a meta-skill for Hermes Agent that automates the creation of reusable skills by monitoring user workflows. It functions by tracking session patterns in the background and proposing structured skill definitions when it detects repeatable sequences. Users can generate both SKILL.md documentation and Python plugin scaffolds directly from these proposals via slash commands or natural language. This tool streamlines the expansion of the Hermes ecosystem by turning manual environment setups and development tasks into permanent, executable assets.", "highlights": [ "Automatically detects and proposes reusable skills from active workflows", "Generates both SKILL.md documentation and Python plugin scaffolds", "Provides slash commands to manage, list, and save detected patterns" ], "readmeHash": "6be34c4d43dbe824a02a1b76e14778d8c7e7dc33e27bda0754918cf72ca17dc0", "generatedAt": "2026-04-17T18:15:58.005Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:01:35.361Z" }, "ReinaMacCredy/maestro": { "summary": "Maestro is a local-first coordination system for human and agent work, written in TypeScript and run with Bun. It stores work, decisions, sessions, evidence, dispatches, and handbacks in each repository’s shared Git root, and provides a mechanism kernel, optional policy plugins, and Markdown recipes and skills.", "highlights": [ "Does not require a background service.", "Includes an SLP protocol for supervised teams.", "Provides CLI commands for installation, updates, and diagnostics." ], "readmeHash": "e946cc7e08b8d847dd4a14851a010543f3c5397f75df98ab1020e0ae54efcf76", "generatedAt": "2026-10-19T15:06:22.523Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "tiann/execplan-skill": { "summary": "ExecPlan is an Agent Skill designed to help AI coding agents autonomously manage complex, long-running implementation tasks. It utilizes a structured execution plan that allows users to review and approve the approach before implementation starts. The system tracks progress in real-time and provides rollback procedures for safe recovery. It is compatible with tools such as Claude Code, Cursor, and Goose.", "highlights": [ "Autonomous execution for long-running tasks", "Real-time tracking of progress and decisions", "Built-in rollback procedures for safe recovery" ], "readmeHash": "f53fd8bd2f029b396c4015da51c28d75ec3566895f270f59edf369ea91e0533f", "generatedAt": "2026-04-17T18:16:08.358Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:01:41.840Z" }, "cablate/Agentic-MCP-Skill": { "summary": "Agentic-MCP-Skill is an experimental MCP client designed to reduce token consumption through a three-layer progressive disclosure pattern. It implements a socket-based daemon architecture that allows AI agents to load server metadata, tool lists, and specific tool schemas incrementally rather than all at once. This approach aims to optimize context window usage and improve the efficiency of managing multiple MCP servers.", "highlights": [ "Three-layer progressive loading to minimize token usage", "Socket-based daemon architecture for persistent connections", "Incremental discovery of server metadata and tool schemas" ], "readmeHash": "f560b32f0477897c25ed2422c5e5e20b0785016e3f5d780725780164f300bdc2", "generatedAt": "2026-07-14T20:11:12.758Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:27:37.781Z" }, "armelhbobdad/bmad-module-skill-forge": { "summary": "Skill Forge analyzes code repositories, documentation, and developer discourse to create instruction files for AI agents. Its workflows can produce skills with source citations or documentation URLs, test them, and export them.", "highlights": [ "Creates agent instructions from repositories, docs, and developer discourse", "Supports automated and step-by-step skill workflows", "Includes skill testing and export" ], "readmeHash": "5c8538b337072165b1f9a21102ea660bfad25110c8e49c2e3d01d73ec35573f5", "generatedAt": "2026-10-12T17:06:48.639Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:07.483Z" }, "amanning3390/hermeshub": { "summary": "HermesHub is an ARD-compliant agent registry that enables the discovery and indexing of AI agent capabilities. It utilizes a Hermes Agent to autonomously crawl, validate, and index manifests hosted at /.well-known/ai-catalog.json endpoints. Users can list agents for free via self-publishing or through a paid hosted listing that includes health checks. The platform supports ranked discovery via a standardized search API and maintains federation with Hugging Face Discover and GitHub Agent Finder.", "highlights": [ "Full implementation of Agentic Resource Discovery (ARD) v0.9 specification", "Autonomous 24/7 operation via Hermes Agent on NVIDIA infrastructure", "Federated search integration with GitHub and Hugging Face" ], "readmeHash": "4092220ff7eebd16b1f0efbcab0d4c5571ccfe1169e83aa7703dc57706c6dbe4", "generatedAt": "2026-07-01T07:57:11.286Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:01:52.554Z" }, "PederHP/skillsdotnet": { "summary": "SkillsDotNet is a C#/.NET implementation of the agentskills.io standard designed to integrate agent skills into the Model Context Protocol (MCP). It uses a skill:// URI convention to implement progressive disclosure, allowing clients to discover lightweight skill metadata before loading full documentation or supporting files. The library provides seamless integration with Microsoft.Extensions.AI and supports automatic discovery of skills from various agent directories like Claude, Cursor, and Copilot.", "highlights": [ "Implements progressive disclosure via SKILL.md files", "Integrates with Microsoft.Extensions.AI for tool-based skill loading", "Supports automatic registration of common agent skill directories" ], "readmeHash": "33b9ac31521386ecd1c35e8b54c624518129069059c39e4e23ac672a590aacde", "generatedAt": "2026-05-10T07:19:07.898Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:01:55.740Z" }, "mem0ai/mem0": { "summary": "Mem0 is a memory layer for AI assistants and agents that stores user, session, and agent state to support personalized interactions over time. It is available as a Python and npm library, a self-hosted server, or a managed cloud platform.", "highlights": [ "Supports semantic, keyword, and entity-based retrieval.", "Offers SDKs, a CLI, self-hosted server, and cloud platform.", "Use cases include assistants, customer support, healthcare, productivity, and gaming." ], "readmeHash": "f72db3426438dbdf6879a2b50583fbee8e7e2dde5d01308c6295bcd09fd3db86", "generatedAt": "2026-10-15T04:55:32.856Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:08.097Z" }, "vectorize-io/hindsight": { "summary": "Hindsight is an agent memory system designed to help agents learn over time, rather than only recall conversation history. It provides retain, recall, and reflect operations, and can be self-hosted or used through Hindsight Cloud.", "highlights": [ "Includes retain, recall, and reflect operations.", "Offers Docker, pip, and Kubernetes deployment options.", "Supports hosted and local LLM providers." ], "readmeHash": "a3e2745dedec4f9e0d4452857ef054699e60414740f5d8f12d41bd42711d476a", "generatedAt": "2026-10-19T15:06:24.394Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "greyhaven-ai/autocontext": { "summary": "autocontext is a harness for improving agent performance through evaluated task runs. It retains useful lessons and produces traces, reports, playbooks, datasets, and optional local-model training artifacts for later runs.", "highlights": [ "Start runs from a plain-language goal or saved scenario.", "Inspect, diff, replay, or export filesystem-based run artifacts.", "Supports multiple agent entry points, including Pi and MCP clients." ], "readmeHash": "726c0568378bf79a78603afeed7777269814b4ad99b828e803483c85d21fa972", "generatedAt": "2026-10-24T04:50:19.462Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "elkimek/honcho-self-hosted": { "summary": "This project provides a configuration layer for self-hosting Plastic Labs' Honcho memory backend, enabling cross-session persistence for Hermes Agent without relying on third-party cloud storage. It deploys the full Honcho stack—including the API, PostgreSQL with pgvector, and Redis—on local hardware using Docker Compose. Users can route memory extraction and reasoning tasks through any OpenAI-compatible API or local inference servers like Ollama to maintain data sovereignty. This setup allows Hermes to build long-term user profiles and logical observations while keeping conversation data on the user's own infrastructure.", "highlights": [ "Enables private cross-session memory for Hermes Agent via self-hosting", "Supports any OpenAI-compatible LLM provider or local inference servers", "Includes automated setup script for Docker, PostgreSQL, and Redis" ], "readmeHash": "97ef52e93810f102b048d8f240e58e6c21d153cd82f14abcca6d545c6f0782b7", "generatedAt": "2026-04-17T18:16:39.713Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:02:08.398Z" }, "yoloshii/ClawMem": { "summary": "ClawMem is an on-device memory layer for Claude Code, OpenClaw, Hermes, and other AI agents, with retrieval-augmented search, hooks, and an MCP server. It uses a local SQLite vault and supports importing conversation exports from Claude Code, ChatGPT, Claude.ai, Slack, and plain text.", "highlights": [ "Combines BM25, vector search, rank fusion, and cross-encoder reranking.", "Generates session handoffs and captures decisions and preferences.", "Integrates through hooks, MCP, and native agent plugins." ], "readmeHash": "0a3303f5bcd6266e4c0d976bb358a7143f85054e326bcf84f40f9b9a49f5aff7", "generatedAt": "2026-10-12T17:06:53.295Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:11.823Z" }, "plur-ai/plur": { "summary": "PLUR provides persistent, shared memory for AI agents through plain-text YAML engrams that users can inspect, correct, and delete. It runs locally and integrates with MCP tools, including Claude Code, Codex, and Cursor.", "highlights": [ "Stores agent memories as plain YAML files.", "Supports local search, with keyword search as a fallback.", "Provides CLI setup and an MCP server." ], "readmeHash": "505128e9cf62ae1f86ada2ef3c7347a746e1fac1cbf79f372fe90fba4276e102", "generatedAt": "2026-10-22T21:48:07.098Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "amanning3390/flowstate-qmd": { "summary": "FlowState-QMD is a local-first memory layer designed to provide coding agents with immediate project context from markdown documentation. It utilizes a SQLite-based vector store and an anticipatory cache that prefetches relevant documents by monitoring active agent session logs. This approach allows agents to access design docs, ADRs, and changelogs before initiating a reactive search loop. The system integrates via an MCP server and supports local execution using GGUF models and node-llama-cpp.", "highlights": [ "Anticipatory cache prefetches context by watching agent session logs", "Local-first storage using SQLite-vec and GGUF models via node-llama-cpp", "MCP-native server with automated configuration for major coding agent clients" ], "readmeHash": "7ec46f4f7fd14b7dc30472f3c4ba0d7c03b2f8a298928edb002ec61cb3ba071c", "generatedAt": "2026-04-21T07:08:50.577Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:02:19.249Z" }, "robbyczgw-cla/hermes-web-search-plus": { "summary": "Web Search Plus is a Hermes plugin that adds tools for web search and clean page extraction. It connects multiple search and extraction providers, can route queries automatically, and can try another provider when one fails or returns no results.", "highlights": [ "Adds `web_search_plus` and `web_extract_plus` tools.", "Supports automatic provider routing and fallback.", "Works with one configured provider." ], "readmeHash": "e66d6725bda52366cff1fd62107ed0142eafab391c9ddfa2d34952355a420a82", "generatedAt": "2026-10-25T22:50:06.510Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "42-evey/hermes-plugins": { "summary": "hermes-plugins is a collection of plugins and workflow skills for Hermes Agent. The plugins cover goals and planning, delegation, safety, memory, observability, and integrations; skills provide guidance for software development and autonomous-agent workflows.", "highlights": [ "Plugins install individually and remain disabled until enabled.", "Plugins use the configured model through Hermes’s ctx.llm.", "The README lists six workflow skills." ], "readmeHash": "36df3f2bfc88a60ea7565f55a7518b8a4444bbe67987a71fda9ef264813a9f9d", "generatedAt": "2026-10-24T18:21:25.643Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "FahrenheitResearch/hermes-weather-plugin": { "summary": "The Hermes Weather Plugin provides a suite of 13 meteorological tools for the Hermes Agent ecosystem, enabling real-time access to observations, forecasts, and high-fidelity model imagery. It utilizes a Python-first architecture that offloads heavy computational tasks, such as NEXRAD radar rendering and ECAPE calculations, to high-performance Rust backends. The plugin features an automated bootstrap process that compiles necessary Rust binaries on first use, ensuring a streamlined setup for complex weather analysis. Users can retrieve verified model soundings, severe weather outlooks from the SPC, and global data via Open-Meteo integration.", "highlights": [ "Thirteen tools for NWS forecasts, radar imagery, and severe weather alerts", "Rust-backed meteorological calculations for ECAPE, soundings, and parcel profiles", "Automated bootstrap system for compiling and caching required Rust binaries" ], "readmeHash": "869d310335a0ee941fcdea6d5d8ec80a4d8bd716de3cfdd2c26f38da8d507a25", "generatedAt": "2026-04-17T18:17:05.522Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:02:28.746Z" }, "42-evey/evey-bridge-plugin": { "summary": "The Evey Bridge Plugin is a technical integration that connects Claude Code with the Evey hermes-agent stack. It functions by automatically checking for messages or tasks from the bridge and injecting them as context into every prompt submission. The plugin includes specialized commands for running a 'Mother Mode' improvement loop and monitoring system health, including wallet and cron status. It is designed specifically for users running the Evey stack at a designated local directory.", "highlights": [ "Injects Evey messages and tasks into Claude Code prompts", "Automates Mother Mode improvement loops via the /evey-loop command", "Monitors health, wallets, and crons with quick status checks" ], "readmeHash": "60bb4ab4be41123be5f6888b661fbf14d036963192420434acca3245a5289d1d", "generatedAt": "2026-04-17T18:17:09.686Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:02:31.642Z" }, "nativ3ai/hermes-payguard": { "summary": "Hermes PayGuard is a security-focused plugin that enables Hermes Agents to handle USDC and x402 payments through a gated execution model. It separates payment preparation from execution by requiring an out-of-band human approval stamp for transfers exceeding user-defined policy limits. The system supports Circle developer-controlled wallets, cross-chain CCTP transfers, and automated x402 micropayments for paid HTTP fetches. This architecture ensures that while agents can stage financial intents, they cannot unilaterally move significant funds without explicit operator consent.", "highlights": [ "Enforces human-in-the-loop approval for high-value USDC transfers", "Supports Circle CCTP for cross-chain USDC routing and execution", "Automates x402 micropayments below configurable policy thresholds" ], "readmeHash": "3009c6924fcf83db62ff7e2507eac8a084268e497c40b19b2c6d42bb5e67ded5", "generatedAt": "2026-04-17T18:17:13.852Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:02:34.575Z" }, "builderz-labs/mission-control": { "summary": "Mission Control is a self-hosted, alpha-stage control plane for operating and governing AI agents. It provides a local dashboard backed by SQLite to dispatch tasks, track token spend, and coordinate agent runtimes such as Claude Code and OpenClaw.", "highlights": [ "Governs tasks, agent registration, memory browsing, and cost tracking.", "Provides Web UI, CLI, MCP server, and REST API interfaces.", "Supports local installation via Node.js or deployment using Docker." ], "readmeHash": "1cfb5a9b7bbfe40280e165fd3ca374103ecd40d3050c9be81917d9e0311e4cd1", "generatedAt": "2026-08-25T07:12:33.323Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-31T15:33:39.248Z" }, "swarmclawai/swarmclaw": { "summary": "SwarmClaw is a self-hosted runtime and framework designed for managing autonomous AI agent swarms and orchestrators. It utilizes a multi-agent architecture featuring delegation, scheduling, and durable memory to coordinate complex workflows. The platform supports over 23 LLM providers and integrates with MCP tools and various CLI backends. It serves as a centralized dashboard for visualizing agent teams and managing multi-agent AI operations.", "highlights": [ "Supports 23+ LLM providers including OpenAI and Ollama", "Features durable agent memory and MCP tool integration", "Provides visual org charts for agent team delegation" ], "readmeHash": "250e36623ed24c8013b6e8ee00707cf8b5d35267e3750648e6d02e2228093cfa", "generatedAt": "2026-07-14T20:13:50.601Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:27:52.550Z" }, "howdymary/hermes-agent-metaharness": { "summary": "Hermes Agent Meta-Harness is a standalone outer-loop optimizer designed to improve LLM system quality by optimizing the code that manages context and retrieval. It functions by treating the Hermes Agent as an execution backend and performing structured searches and mutations over harness candidates. The tool facilitates automated benchmark evaluation, baseline comparison, and frontier tracking for verifiable coding tasks. It is specifically configured to target benchmarks like TBLite and TerminalBench-2.", "highlights": [ "Automates outer-loop optimization of benchmark harness code", "Orchestrates TBLite and TerminalBench-2 benchmark evaluations", "Provides structured candidate mutation and frontier tracking" ], "readmeHash": "63589bd1e2faacb5f76d6531a3b50e24376a1699a104358b933848bcd3071735", "generatedAt": "2026-07-14T20:14:01.285Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:27:51.879Z" }, "runtimenoteslabs/gladiator": { "summary": "Gladiator is an autonomous agent arena that pits two AI companies against each other to compete for GitHub stars on identical repositories. The system uses Paperclip for orchestration and Hermes Agent workers to execute tasks according to distinct corporate strategies: growth-focused Blitz versus quality-focused Craft. A real-time dashboard visualizes the competition through scoreboards, task boards, and audit trails. Upon completion, the project demonstrates cross-company skill transfer by merging the rival entities and their learned memories.", "highlights": [ "Orchestrates 9 autonomous agents across two competing AI companies", "Uses Paperclip and Hermes Agent for task execution and learning", "Real-time dashboard visualizes agent tasks, code, and scoreboard metrics" ], "readmeHash": "d2f920c1ac8439bf967a4733beffdf47f47c7516ad6f85a89ffa08271c09bcbe", "generatedAt": "2026-04-17T18:17:34.239Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:02:53.131Z" }, "1ilkhamov/opencode-hermes-multiagent": { "summary": "Hermes is a multi-agent orchestration system for OpenCode AI that coordinates 17 specialized agents to handle software development tasks. The system uses a pipeline architecture to manage processes such as research, planning, implementation, quality, documentation, and infrastructure.", "highlights": [ "Coordinates 17 specialized AI agents", "Uses a strict pipeline architecture", "Includes a Master Orchestrator for routing and coordination" ], "readmeHash": "b32708f1fdf20bbf066766e0d21ce94c9e792db0221c76e44c14f4cf1f5b9595", "generatedAt": "2026-07-14T21:30:58.742Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:37:08.129Z" }, "numtide/llm-agents.nix": { "summary": "llm-agents.nix provides Nix packages for AI coding agents and development tools, with packages automatically updated daily. The README lists tools including terminal coding agents, desktop applications, and development platforms, with example `nix run` commands for each.", "highlights": [ "Packages AI coding agents and development tools for Nix.", "Packages are automatically updated daily.", "README provides `nix run` examples for listed tools." ], "readmeHash": "571f8d519414fb8cc64c78f6c6201f2f8d368d2a2bf3f5d6a18e6a4b9c458ca4", "generatedAt": "2026-10-25T16:35:33.800Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "TheAiSingularity/hermesclaw": { "summary": "HermesClaw is a community implementation that runs the Hermes Agent inside an NVIDIA OpenShell sandbox to provide hardware-level security enforcement. It utilizes kernel-level mechanisms like Landlock LSM and Seccomp BPF to restrict network egress, filesystem access, and dangerous system calls. This architecture allows the agent to maintain its full capability stack while ensuring that rogue skills are stopped by the operating system. Users can manage security postures through policy presets ranging from strict inference-only to permissive web and messaging access.", "highlights": [ "Hardware-enforced sandboxing via NVIDIA OpenShell", "Kernel-level restriction of network and filesystem access", "Configurable security policies for messaging and web search" ], "readmeHash": "0a892ee229ccdb852dcebfa8fe0b0327615f72db3e40cae535ad085ebf603cf9", "generatedAt": "2026-07-14T21:04:37.509Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:27:53.611Z" }, "0xrsydn/nix-hermes-agent": { "summary": "nix-hermes-agent provides a Nix package and NixOS module for the declarative deployment of Hermes Agent. It allows users to manage the entire agent lifecycle—including configurations, workspace documents, and systemd services—through a single Nix configuration file. The module features a robust update policy that uses quarantined candidate pull requests to prevent broken upstream releases from affecting stable deployments. It also supports secure secrets management by keeping API keys and authentication tokens outside of the Nix store.", "highlights": [ "Declarative configuration for cli-config.yaml and workspace documents", "Hybrid skill management supporting both Nix-managed and CLI-installed skills", "Secure secrets integration via systemd EnvironmentFiles and sops-nix" ], "readmeHash": "545cb9e371fb591a0e1d8cf3b96c5af99ec688f3bd1371df443485644ce14b12", "generatedAt": "2026-07-14T21:06:18.900Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:27:54.620Z" }, "xmbshwll/hermes-agent-docker": { "summary": "This project provides a minimal Docker sandbox image designed to package and deploy Hermes Agent. It utilizes the official upstream installation script to set up the agent and includes mini-swe-agent within the container. The image supports multi-arch publishing and allows for persistent state management by mounting the agent's home directory. Users can easily build specific versions using build arguments to target different Hermes branches or tags.", "highlights": [ "Includes mini-swe-agent and official Hermes installation", "Supports persistent state via mounted home directories", "Allows building specific Hermes tags using build arguments" ], "readmeHash": "d4df3d5de4e1cbe035758f17ef87c3f31e85c8eb025649c0daf4c2f9db340cc1", "generatedAt": "2026-04-17T18:18:11.811Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:04:07.849Z" }, "JackTheGit/hermes-autonomous-server": { "summary": "This project provides a deployment framework for running Hermes Agent as a persistent, headless Linux service. It utilizes systemd to manage the Hermes Gateway, ensuring the agent remains active across reboots without requiring an open terminal session. The setup integrates with the Nous Portal for model access and leverages the native Hermes cron scheduler to automate recurring AI tasks. This approach offers a stable, self-hosted alternative to manual shell loops or third-party hosting for autonomous server operations.", "highlights": [ "Configures Hermes Gateway as a reboot-safe systemd service", "Enables autonomous execution of recurring AI cron jobs", "Supports headless operation on Ubuntu and Debian servers" ], "readmeHash": "20f77b85b2db065f9c329bfc6a0750d016c1c269ced5ce3cdbcdf09466720073", "generatedAt": "2026-04-17T18:18:15.996Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:04:11.181Z" }, "teknium1/hermes-miniverse": { "summary": "Hermes-miniverse is a bridge that enables Hermes Agents to inhabit and interact within Miniverse pixel worlds. It functions through a three-part architecture consisting of a standalone bridge daemon, a gateway hook for state broadcasting, and a specific skill that allows the agent to consciously navigate the environment. This integration allows agents to maintain a visual presence, update their status in real-time, and engage in bidirectional communication with other AI entities. Users can deploy multiple bridges simultaneously to populate a virtual world with specialized Hermes instances.", "highlights": [ "Connects Hermes Agents to Miniverse pixel-based virtual environments", "Synchronizes live agent states including working, thinking, and idle", "Enables bidirectional messaging between Hermes and other world agents" ], "readmeHash": "fe973d184c22ca8e195604ee6bc90547c24992b500537a71fe2f4d05a2f87d88", "generatedAt": "2026-04-17T18:18:28.234Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:04:32.435Z" }, "Ridwannurudeen/hermes-council": { "summary": "Hermes Council is an MCP server designed to provide an adversarial judgment layer for Hermes Agent to prevent overconfident or unsafe autonomous actions. It functions by employing multiple specialized personas—such as Advocate, Skeptic, and Arbiter—to deliberate on plans, code diffs, and claims. This structured deliberation produces explicit verdicts including risk assessments, required checks, and verified evidence. The toolset also supports RL workflows by extracting DPO preference pairs and normalized rewards.", "highlights": [ "Provides adversarial preflight reviews for high-stakes autonomous actions", "Stress-tests implementation plans and code diffs for potential failures", "Generates structured verdicts and DPO preference pairs for RL training" ], "readmeHash": "ffb631ba0a5cabcb024d99713797c55fe6d35ece32ed598be9ffc4fcc6d25c58", "generatedAt": "2026-07-14T21:06:08.411Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:27:57.132Z" }, "gizdusum/hermes-blockchain-oracle": { "summary": "Hermes Blockchain Oracle is a Model Context Protocol (MCP) server that provides Hermes Agent with real-time access to the Solana blockchain. It functions as a bridge between natural language queries and on-chain data by exposing specialized tools for wallet, token, and transaction analysis. The server enables users to monitor network health, track whale movements, and manage NFT portfolios through conversational interactions. By integrating directly with Solana RPC nodes, it eliminates the need for manual block explorer searches.", "highlights": [ "Provides real-time Solana wallet balances and token portfolio tracking", "Detects large-scale whale transfers with configurable movement thresholds", "Retrieves NFT metadata, collection details, and floor price information" ], "readmeHash": "945a1443f02bfbfec395e29e7a946705c8148e9f3da794a013503db9dedf354d", "generatedAt": "2026-04-17T18:18:42.037Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:04:39.116Z" }, "junhoyeo/tokscale": { "summary": "Tokscale is a CLI tool and visualization dashboard for monitoring and analyzing token consumption and costs across multiple AI coding agents. Its README lists support for agents including Claude Code, OpenClaw, Codex CLI, Gemini CLI, and Hermes Agent.", "highlights": [ "Tracks token usage and costs across multiple AI coding agents.", "Provides a CLI and visualization dashboard.", "Supports a range of coding agents and tools." ], "readmeHash": "d5796fc210232fef3d4739e59a200caa3717835ab9d0817b26ff328101ba5645", "generatedAt": "2026-10-25T05:17:39.364Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "joeynyc/hermes-skins": { "summary": "Hermes Skins is a collection of custom visual themes for the Hermes CLI agent. Skins can change colors, spinner text, labels, branding, and ASCII art without changing the agent’s personality or behavior; users can install YAML files and activate them for a session or in their configuration.", "highlights": [ "Includes custom and built-in skin listings.", "Custom skins inherit unspecified values from the default skin.", "Users can create skins with YAML files." ], "readmeHash": "5e35608cff3608b6f85830bb4ba604551b06bb8a2670d78589a0c52d50b39de0", "generatedAt": "2026-10-22T12:09:42.051Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "AlexAI-MCP/hermes-CCC": { "summary": "hermes-CCC is a port of the Hermes Agent framework specifically designed for the Claude Code native skill system. It integrates the Hermes brain architecture directly into the Claude Code execution context, eliminating the need for external processes or OAuth authentication. The project implements 46 core skills across categories such as MLOps, software development, and research, including specialized tools for RL training and vector database management. This allows developers to utilize Hermes-style memory, personas, and task routing while remaining within a unified terminal environment.", "highlights": [ "Port of Hermes Agent brain and 46 core skills to CCC", "Runs natively in Claude Code without separate processes or OAuth", "Includes MLOps tools for vLLM, GGUF inference, and RL training" ], "readmeHash": "7a7321cf0c74bea4af5263995a5d927c7027d6ea5aa08c8dbe4b7e3076db7946", "generatedAt": "2026-04-17T18:18:54.559Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:04:49.059Z" }, "unmodeled-tyler/vessel-browser": { "summary": "Vessel is an open-source web browser built on Chromium that integrates a sidebar AI assistant for tasks such as research, form filling, and page navigation. It features a supervisor interface that allows users to monitor real-time AI actions, approve sensitive operations, and undo recent changes.", "highlights": [ "Built-in support for Anthropic, OpenAI, Ollama, and other AI providers.", "Supervisor sidebar for managing agent approvals, checkpoints, and conversation threads.", "Supports external MCP clients and browser automation workflows." ], "readmeHash": "9cd764a28b255f857958078722f4f290347770b80e4b039d513619124278e038", "generatedAt": "2026-09-15T05:03:38.761Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:36:08.650Z" }, "Abruptive/Ankh.md": { "summary": "Ankh.md is an open-source framework designed to transform a single Hermes Agent into multiple specialized, project-specific instances. It works by detecting a local `.agent/` directory and dynamically overriding the default `hermes` command with a scoped agent configuration. This allows developers to maintain unique instructions, skills, and memories for different codebases or workflows without cross-project interference.", "highlights": [ "Per-folder scoping with local configuration and skills", "Seamlessly switches between default and scoped Hermes agents", "Supports project-local memories, tools, and custom instructions" ], "readmeHash": "dd4b8e378f4fca75140a50a8fb1709c3572c7e801ded42ea75e935366c029378", "generatedAt": "2026-04-17T18:19:04.079Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:04:55.412Z" }, "0xNyk/openclaw-to-hermes": { "summary": "openclaw-to-hermes is a Python-based migration tool designed to transfer agents, workspaces, and configurations from OpenClaw to Hermes Agent. It maps model settings, credentials, and markdown-based memories while creating backups and supporting dry-run previews.", "highlights": [ "Migrates SOUL.md, MEMORY.md, skills, and memory archives to Hermes.", "Maps OpenClaw configs and credentials into Hermes YAML and JSON formats.", "Sets up systemd gateway services and handles nested auth structures." ], "readmeHash": "5f13e185b327e4859a5883e9af58379ff7fad3539bc3bf392d6b72410250fed1", "generatedAt": "2026-07-17T18:41:42.973Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-20T08:59:28.945Z" }, "42-evey/evey-setup": { "summary": "Evey Setup is a bootstrapping tool designed to quickly deploy a complete autonomous AI agent stack based on hermes-agent. It uses a series of bash scripts to automate the installation of Docker containers, model routing, and community plugins. The system integrates a variety of services including LiteLLM for model proxying, Ollama for local inference, and Qdrant for vector memory. Users can choose between three deployment tiers—Base, Services, or Full—to balance resource usage with available functionality.", "highlights": [ "Automated installation of hermes-agent and 12+ supporting services", "Three scalable deployment tiers from minimum viable to production stack", "Support for 29 community plugins across six functional categories" ], "readmeHash": "bd2f9c01f6716a864e01cf898abfb435bcd96c10b9d17894eaf6eefd1051e6c4", "generatedAt": "2026-06-29T08:20:11.399Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:05:03.761Z" }, "ucsandman/DashClaw": { "summary": "DashClaw is an approval layer that can intercept tool calls through supported integrations before they execute. It lets users review and resolve pending actions remotely, and records decisions in a ledger.", "highlights": [ "Supported hooks and gateway can cancel blocked calls before execution.", "Pending actions can be resolved through an approvals inbox, CLI, phone PWA, Telegram, or Discord.", "Decision records and signed exports support later review." ], "readmeHash": "711e3a30427219e9537267d5d4b273b16fb5ee7fb8e53c151af7f6d417702adc", "generatedAt": "2026-10-19T15:06:38.926Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Lethe044/hermes-incident-commander": { "summary": "Hermes Incident Commander is an autonomous SRE agent designed to detect, diagnose, and remediate production infrastructure incidents. It operates either as a plugin for Hermes Agent or as a standalone watchdog on Linux hosts using an Anthropic API key to provide triage and automated healing.", "highlights": [ "Monitors CPU, memory, disk, and failed systemd units.", "Supports automated remediation and adaptive, time-of-day-aware thresholds.", "Integrates with PagerDuty, Slack, Discord, and Microsoft Teams." ], "readmeHash": "094171f5d6e613df485aa2c69682d32cdefaedc3770be2b996124e8b4903f0c5", "generatedAt": "2026-09-28T16:11:13.787Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass" }, "Christabel337/job-scout-agent": { "summary": "Job Scout Agent is an autonomous tool that fetches job listings from multiple APIs and scores them against user skills and preferences. It performs company research and saves results into organized folders and a CSV tracker.", "highlights": [ "Fetches jobs from live job board APIs", "Scores listings against user skills and preferences", "Updates a CSV tracker for Google Sheets" ], "readmeHash": "1c306991286128e628d02b8f35a0bcd1fccfe8d25a4c32ca3e6d16980af92b47", "generatedAt": "2026-07-14T21:31:18.981Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:37:08.045Z" }, "rodmarkun/anihermes": { "summary": "AniHermes is a natural language media server that simplifies the process of downloading and tracking anime. It leverages the Hermes Agent for LLM-driven tool orchestration to manage library tasks through conversational interfaces on Telegram, Discord, or the CLI. The system integrates with qbittorrent for downloads and Anilist for watchlist synchronization. Users can perform complex actions like identifying anime from screenshots via vision analysis or scheduling automatic weekly downloads.", "highlights": [ "Natural language control via Telegram, Discord, and CLI", "Automated downloads using qbittorrent and Anilist synchronization", "Vision-based anime identification from uploaded screenshots" ], "readmeHash": "a1b733dea4e81c832c651e14526a811c4bed10c584a76df67b98ce8e95e38e20", "generatedAt": "2026-06-15T10:14:45.110Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:05:16.165Z" }, "JackTheGit/hermes-ai-infrastructure-monitoring-toolkit": { "summary": "This project is a reference implementation for building autonomous, long-running monitoring systems using Hermes Agent. It utilizes cron-based workflows to ingest AI research from arXiv and track infrastructure repositories on GitHub. The system automatically generates structured markdown reports covering compute risk analysis, cost projections, and trend dashboards. It serves as a template for developers to deploy headless, scheduled agents that transform raw research data into actionable infrastructure intelligence.", "highlights": [ "Automated research ingestion via arXiv API and GitHub ecosystem tracking", "Scheduled cron workflows for risk analysis and cost modeling", "Generates terminal-style dashboards and structured markdown intelligence reports" ], "readmeHash": "503ac8269ccf34295975dda8ba840a2d727201aa7cc30f9497ebea699e06a9b7", "generatedAt": "2026-05-04T11:34:23.373Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:05:18.814Z" }, "bigph00t/hermescraft": { "summary": "HermesCraft is an embodied AI framework that integrates Hermes agents into Minecraft as persistent, interactive players. It connects the Hermes agent stack to a Mineflayer bot body through a custom Minecraft CLI and HTTP API, allowing agents to perceive and interact with the game world. Agents can function as individual companions or as part of multi-agent civilizations with private messaging and social memory. The system emphasizes fair-play perception by restricting agent knowledge to line-of-sight visuals and directional audio rather than using privileged game data.", "highlights": [ "Embodied agents with persistent memory, identity, and tool-use capabilities", "Fair-play perception system using line-of-sight filtering and spatial mapping", "Multi-agent social features including public chat and private messaging" ], "readmeHash": "a1fc221856f4d660b6ab467e7195032aa74651d343d397d395308bc2c33808d8", "generatedAt": "2026-04-17T18:19:45.708Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:05:21.782Z" }, "hxsteric/mercury": { "summary": "Mercury is a blockchain cash flow analyzer and Hermes Agent skill designed to provide transparency into wallet activities across Ethereum, Base, and Solana. It utilizes parallel subagents to fetch transaction data, identify six specific fraud patterns, and generate AI-driven trading strategies based on wallet behavior. The project features a multi-layer WebGL and Canvas dashboard for real-time visualization of inflows, outflows, and risk metrics. It operates as a standalone tool or an integrated Hermes skill for automated multi-chain investigations.", "highlights": [ "Analyzes transaction flows across Ethereum, Base, and Solana chains", "Detects six fraud patterns including mixer interactions and dust attacks", "Visualizes data via interactive WebGL and Canvas-based dashboards" ], "readmeHash": "59c939a620fea0efb540523db902d95e5842e9f9a587ec97b6bc26d3d9970587", "generatedAt": "2026-04-17T18:19:49.582Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:05:24.528Z" }, "bryercowan/hermes-embodied": { "summary": "Hermes Embodied is a robotics extension for Hermes Agent that enables autonomous fine-tuning of Vision-Language-Action (VLA) models through natural language. The system orchestrates a self-improvement loop by collecting robot trajectories in simulation or hardware, provisioning cloud GPUs via Vast.ai, and retraining models when performance thresholds are met. It integrates with the LeRobot framework and MuJoCo simulation to automate the transition from data collection to model promotion. This allows users to manage complex robotics training pipelines without manual machine learning expertise.", "highlights": [ "Automates VLA fine-tuning loops using natural language commands", "Provisions and manages Vast.ai cloud GPU instances autonomously", "Supports SmolVLA and GR00T models for simulation or physical hardware" ], "readmeHash": "3a942dc575c8725fdeff8c20356ae232bcb780516b5eb988d6a11c58e0d82766", "generatedAt": "2026-04-17T18:19:53.722Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:05:27.308Z" }, "0xNyk/awesome-hermes-agent": { "summary": "Awesome Hermes Agent is an independent directory of skills, tools, and resources for the Nous Research self-improving agent. It categorizes community-contributed plugins, memory providers, and integrations using editorial maturity tags ranging from experimental to production.", "highlights": [ "Directory for skills, plugins, and memory providers", "Categorizes entries by production, beta, and experimental status", "Includes links to official Nous Research core repositories" ], "readmeHash": "8ac0e147d16cd13f62d8e81ff35d285f0d5b13b0a30b506540b93bd9acfb74cb", "generatedAt": "2026-09-12T11:04:42.694Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-14T14:28:38.458Z" }, "alchaincyf/hermes-agent-orange-book": { "summary": "Hermes Agent 2.0: The Complete Guide is a practical guide to the Hermes Agent framework, published as part of the Orange Book series. It covers self-improvement, memory, skills, integrations, multi-agent orchestration, deployment, and security across 21 sections in six parts, with English and Chinese PDF editions available.", "highlights": [ "Covers Hermes Agent across 21 sections and six parts", "Includes English and Chinese PDF editions", "Released under the MIT License" ], "readmeHash": "79cf0ac4b598fcd4ec6bed1ecbb4ceb2e226e1e07175d0047bae9621e8eecbe4", "generatedAt": "2026-10-20T11:51:11.267Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "OnlyTerp/hermes-optimization-guide": { "summary": "A practical guide to configuring, using, and maintaining Hermes Agent, with advice on models, token costs, memory, security, messaging, automation, and troubleshooting. It includes measured token-cost examples and says its commands and configuration references are checked against Hermes Agent v0.21.4 by CI.", "highlights": [ "Covers setup, optimization, security, and production use.", "Includes a troubleshooting chapter and practical recipes.", "CI checks references against Hermes Agent v0.21.4." ], "readmeHash": "6630042717bf51d4dfedfc4f0e5c1e2da11ceafc8565925465af3bc41f0bfd3d", "generatedAt": "2026-10-19T15:06:40.803Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "metantonio/hermes-wsl-ubuntu": { "summary": "This project provides a comprehensive guide and automation scripts for deploying the Hermes Agent with llama.cpp and Qwen3.5 on local environments. It facilitates the end-to-end setup of an AI agent stack including GPU acceleration via CUDA for Linux/WSL2 or Metal for macOS. The workflow integrates Camofox for browser automation and Hermes HUDUI for a graphical web interface.", "highlights": [ "Automated installation scripts for Ubuntu/WSL2 and macOS", "Supports GPU acceleration via CUDA and Metal backends", "Integrates Camofox browser server and Hermes HUDUI" ], "readmeHash": "1a8a5893c46daa39adbf212f4189d3c9c007157deb2367e25cb9e40ee8242f04", "generatedAt": "2026-07-14T21:06:39.687Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:28:03.642Z" }, "mudrii/hermes-agent-docs": { "summary": "Hermes Agent is an autonomous AI agent developed by Nous Research that features a built-in learning loop for skill creation and memory persistence. It supports multi-platform messaging, various model providers, and an extensible plugin architecture.", "highlights": [ "Creates and improves reusable skills from experience", "Supports multiple messaging gateways and IDE integrations", "Features a pluggable architecture for custom tools and providers" ], "readmeHash": "fc0130614bd2a1b64c189bf3727f5753bb8ac77d29552f958cffa50b361533f3", "generatedAt": "2026-07-14T21:31:26.387Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:37:11.646Z" }, "martymcenroe/HermesWiki": { "summary": "HermesWiki provides documentation for Hermes, an autonomous email agent that triages inbound recruiter outreach. The wiki covers the system's architecture, conversation state machine, guardrails, evaluation, security, and cost engineering.", "highlights": [ "Covers threat modeling against OWASP and MITRE frameworks", "Includes NIST alignment and cost breakdowns" ], "readmeHash": "7caa2ca6034854421966a8ebe6df6ab8eb08c1722b727c5ec96a7dffd00407f1", "generatedAt": "2026-07-16T18:44:05.555Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-16T21:18:45.275Z" }, "nativ3ai/hermes-agent-camel": { "summary": "This project is a self-improving AI agent designed for autonomous execution with integrated safety boundaries. It utilizes a closed learning loop to create and refine skills from experience while employing a CaMeL-style runtime guard to prevent indirect prompt injections. The agent supports a wide array of LLM providers and can be deployed across various environments, including serverless infrastructure and cloud VMs. Users can interact with the agent via a dedicated terminal interface or through messaging platforms like Telegram, Discord, and Slack.", "highlights": [ "Self-improving learning loop for autonomous skill creation and refinement", "CaMeL-style runtime guard to block unauthorized sensitive tool side effects", "Multi-platform access via CLI and various messaging gateway integrations" ], "readmeHash": "e44eba4dae59de6cc756468ddac68e68b46b4c367c9f6798237daa8e437a54d5", "generatedAt": "2026-06-01T09:58:18.365Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:05:51.709Z" }, "kaminocorp/hermes-alpha": { "summary": "Hermes Alpha is an experimental framework designed to test if a stock AI agent can bootstrap an autonomous bug bounty system from first principles. It utilizes a two-agent loop where a persistent Overseer agent builds, deploys, and iteratively improves a tactical Hunter agent tasked with finding software vulnerabilities. The system operates via a FastAPI-powered web terminal and leverages Elephantasm for memory streams to refine the Hunter's skills based on real-world outcomes. This approach tests whether an agent can evolve its own codebase and strategy without pre-built scaffolding or custom tools.", "highlights": [ "Self-improving loop where Overseer evolves Hunter's code and strategy", "Deployed on Fly.io with a FastAPI and xterm.js web terminal", "Integrated with Elephantasm for long-term memory and performance monitoring" ], "readmeHash": "3aea541a15e77030c8871a9071a0aff35aa7d2eb68e3e95d0ab9f39ab00882ea", "generatedAt": "2026-06-08T09:42:58.351Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:05:54.465Z" }, "beardthelion/hermes-skill-distillation": { "summary": "This project provides a skill distillation pipeline designed to fine-tune Hermes 4 models using real-world agentic trajectories. It captures tool-use sessions from a diverse 30-task battery and converts them into SFT-ready data or live RL signals for the Atropos training framework. The system uses a multi-dimensional reward function to score trajectories based on task completion, efficiency, and error recovery. This approach enables a closed learning loop where production-style agent interactions serve as the primary training data for future model iterations.", "highlights": [ "Generates SFT-ready JSONL data from real-world tool usage trajectories", "Integrates with Atropos for live reinforcement learning via GRPO", "Automated scoring based on completion, efficiency, and error recovery" ], "readmeHash": "5d273698fd2125b6e22f85911ef0835d7c21c8e9fe4296870c155393c5dbbaa9", "generatedAt": "2026-05-18T09:11:50.136Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:05:57.192Z" }, "xaspx/hermes-control-interface": { "summary": "Hermes Control Interface is a self-hosted web dashboard designed to centralize the management of the Hermes AI agent stack. It utilizes a Node.js and Express backend with a Vanilla JS frontend to provide a unified interface for agent orchestration and system monitoring. The platform includes specialized tools for MCP server control, token analytics, and a multi-panel swarm monitor for tracking agent pipelines. It also implements role-based access control to isolate agent profiles across different user teams.", "highlights": [ "Real-time chat with streaming and tool call visualization", "MCP server control plane for managing server lifecycles", "ZOO Swarm Monitor for tracking agent health and kanban pipelines" ], "readmeHash": "cef9d3296a8e4f7d74817f63ad25feb2c4c86f7480b85334a9c63f40dfa185a4", "generatedAt": "2026-06-28T07:50:12.557Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:06:00.039Z" }, "ksimback/hermes-ecosystem": { "summary": "Hermes Atlas is a community-curated directory of tools, skills, and integrations for Hermes Agent. It offers categorized project listings with search and filters, GitHub star data, and a chatbot that answers questions using research files.", "highlights": [ "Lists quality-filtered projects across 12 categories.", "Displays live GitHub star counts and growth over time.", "Includes a research-grounded chatbot for Hermes Agent questions." ], "readmeHash": "8a4c4aa462100d165a417de8c6bdddc20f7cbc302e2a02936f4022445a7e7e51", "generatedAt": "2026-10-12T12:36:55.333Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:04:04.942Z" }, "dodo-reach/hermes-desktop": { "summary": "Hermes Desktop is a native macOS companion app that provides a dedicated workbench for managing Hermes Agent without relying on browser wrappers or gateway APIs. It connects directly to a local or remote host via SSH to interact with the real machine-first workflow. The app allows users to manage sessions, Kanban boards, cron jobs, and skills while maintaining the host as the sole source of truth. It is designed to complement the official web dashboard by focusing on direct work and system repair via an embedded terminal.", "highlights": [ "Direct local or SSH connection to Hermes hosts", "Native views for sessions, Kanban, skills, and cron jobs", "Embedded terminal for direct shell access and system repair" ], "readmeHash": "f87a146656f9be7c86a8acfd0be471b824c21d2f195a88904aa74ee99a073fd8", "generatedAt": "2026-06-19T08:21:44.653Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:06:05.716Z" }, "nesquena/hermes-webui": { "summary": "Hermes WebUI is a browser-based interface for Hermes Agent, with chat, session navigation, and workspace file browsing. The README says it uses the existing Hermes agent and models, and provides controls for profiles, settings, and sessions.", "highlights": [ "Three-panel layout for sessions, chat, and workspace files", "Includes a workspace file browser with inline previews", "Designed for access through an SSH tunnel" ], "readmeHash": "4bd873a4b4e7ebd717d9da99c10d34aa1131130bf481f0aa8c6b35a3139a51fa", "generatedAt": "2026-10-23T11:54:53.669Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "garrytan/gbrain": { "summary": "GBrain stores explicit facts with their sources, supports corrections and withdrawal, and makes memory available across agents. It starts with keyless storage and keyword retrieval, with optional semantic search, synthesis, and background enrichment; it can also expose a brain over MCP.", "highlights": [ "Stores facts with provenance and supports corrections and withdrawal.", "Shares memory across agents; transient task state stays local.", "Optional features include semantic search, synthesis, and background enrichment." ], "readmeHash": "0d7e1f97320771a0e1b10d46d27466c2c69841238d0d2bb094c2f918d722a0a6", "generatedAt": "2026-10-18T04:48:44.032Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:28.492Z" }, "pyrate-llama/hermes-ui": { "summary": "Hermes UI serves as a centralized command center for managing and interacting with the Hermes Agent. It operates as a single-file HTML application powered by React 18 and a lightweight Python proxy server. Users can monitor agent activity through real-time log streaming, manage skills, and browse files within a glassmorphic interface. The platform also provides specialized tools for managing MCP servers, cron jobs, and workspace files.", "highlights": [ "Glassmorphic interface for chat and real-time log streaming", "Integrated file browser and task tracking dashboard", "Management tools for skills, cron jobs, and MCP servers" ], "readmeHash": "709ceeaedfd53c3c2bd1d909751f9dc956b3633b3b8c0f2fd43d5393f1f8add2", "generatedAt": "2026-07-01T08:03:25.933Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:06:18.860Z" }, "kevinmarmstrong/hermes-claude-code-rc": { "summary": "This Hermes Agent plugin enables remote management of Claude Code sessions through Telegram or other Hermes-connected platforms. It bridges the gap between local filesystem access and mobile convenience by starting a Claude Remote Control session on a host machine via slash commands. The plugin monitors subprocess output to capture session URLs, allowing users to control their local environment from the Claude mobile app without needing a terminal. It also supports headless command execution and shell access with built-in safety filters for dangerous operations.", "highlights": [ "Starts Claude Code remote sessions via Telegram slash commands", "Provides full local filesystem and tool access from mobile devices", "Executes headless commands and shell operations with safety filtering" ], "readmeHash": "e183d1799625c66d99137ab95f1a26ecc8dfee564cabb5693cf5e552935ae7ca", "generatedAt": "2026-07-06T16:03:52.245Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:06:21.803Z" }, "AaronWong1999/hermesclaw": { "summary": "HermesClaw lets Hermes Agent, OpenClaw, and OpenCode use the same WeChat account. It acts as the sole iLink poller, routing messages to local proxies for Hermes and OpenClaw and an ACP bridge for OpenCode. Users can switch agents with commands such as `/hermes`, `/openclaw`, `/opencode`, `/both`, and `/three`.", "highlights": [ "Routes iLink messages to Hermes, OpenClaw, and OpenCode.", "Supports switching agents with WeChat commands.", "Includes a one-command installer." ], "readmeHash": "977fec254f6a701cab3e6d33e8c77f62d3bfe5c29e1895ec04dd0e2146106f1d", "generatedAt": "2026-10-20T17:22:35.453Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "stephenschoettler/hermes-lcm": { "summary": "hermes-lcm is a context management plugin for Hermes Agent that replaces standard one-shot compression with a SQLite-backed, DAG-based engine. It preserves raw messages and builds depth-aware summary nodes, allowing agents to recover specific details through a suite of recall tools without exceeding active prompt limits.", "highlights": [ "Uses a SQLite store and hierarchical DAG for message persistence.", "Provides recall tools like lcm_grep and lcm_recall for detail recovery.", "Supports Python 3.11+ and integrates as a Hermes Agent plugin." ], "readmeHash": "ec6a734c4f9378664309df4103feb5bbd1bb4b0a45d21320f8dafff4c076775c", "generatedAt": "2026-09-04T10:54:52.820Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-07T13:43:36.362Z" }, "Codename-11/hermes-relay": { "summary": "Hermes-Relay is a native Android companion and CLI tool for the Hermes agent platform. It enables streaming chat, voice interaction, and agent management from mobile devices while the agent's core processing remains on a separate machine.", "highlights": [ "Android app for chat, voice, and agent management.", "Beta CLI for remote file, terminal, and screenshot access.", "Supports standard Dashboard/Gateway connections over LAN or Tailscale." ], "readmeHash": "3658be1f9c85ccdebd3e7d2c6e3da6fcbe89c37cc954be274066664749069341", "generatedAt": "2026-09-03T04:42:10.375Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-07T13:43:36.697Z" }, "clawvader-tech/hermes-telegram-miniapp": { "summary": "Hermes Telegram Mini App is a terminal-style web interface for the Hermes agent that operates within Telegram. It provides a dark-mode dashboard for chatting with agents, managing scheduled cron jobs, and monitoring system resources like CPU and memory.", "highlights": [ "Terminal chat with streaming responses and file attachment support", "Dual HMAC-SHA256 and Ed25519 authentication for Telegram validation", "Background agent spawning with support for five concurrent instances" ], "readmeHash": "c5787ca063daa34c89be1aca78e8d211eb0efb9ab27d09e5757bcbf486dfc8b8", "generatedAt": "2026-07-14T21:31:27.226Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:37:15.900Z" }, "briancaffey/hermes-otel": { "summary": "hermes-otel is an OpenTelemetry plugin for Hermes Agent that turns agent activity—including model calls, tool calls, approvals, and delegated sub-agents—into traces. It records token, cost, and latency metrics, with optional logs and structured events, and exports over OTLP HTTP or gRPC.", "highlights": [ "Traces agent turns, tool calls, approvals, and sub-agents", "Exports over OTLP HTTP or gRPC", "Emits OpenInference and OTel GenAI attributes" ], "readmeHash": "e001bcf0c7fd361568e7edd680692ea07657589c28010fd00d97348bfb0d84b3", "generatedAt": "2026-10-19T15:06:45.657Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "AMAP-ML/SkillClaw": { "summary": "SkillClaw integrates with AI agents to evolve and manage skills from interactions across sessions, agents, devices, and users. It describes merging, deduplicating, improving, and sharing skills, and provides installation instructions for macOS, Linux, and Windows.", "highlights": [ "Integrates with Hermes, Codex, Claude Code, and other agents.", "Unifies skills across agents and devices.", "Offers a dashboard for skills, validation, versions, and session traces." ], "readmeHash": "711f607d6970dd47b2c857127e9cb8c6a66092ab092052f516cb561c3000fba4", "generatedAt": "2026-10-21T05:03:54.674Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "aivrar/portable-hermes-agent": { "summary": "Portable Hermes Agent is a Windows desktop interface for Hermes Agent, with chat, session management, and guided setup. It includes tools for local models through LM Studio, music and speech generation, ComfyUI, workflows, and creating custom tools; setup runs from a folder without administrator rights or Docker.", "highlights": [ "Dark-themed GUI with multilingual support and session management", "Includes workflow automation and runtime tool creation", "Supports local models through LM Studio" ], "readmeHash": "6b2537a1b8eb7852a8f7bb10ec74414d8a9f09a5a679192483c632f3dd6e1c11", "generatedAt": "2026-10-20T17:22:36.381Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Cranot/super-hermes": { "summary": "Super Hermes is a set of skills for the Hermes Agent that enables the agent to generate its own thinking instructions for specific problems. It provides analytical lenses to identify structural trade-offs, bugs, and analysis constraints.", "highlights": [ "Generates custom thinking instructions for specific tasks", "Includes pre-built analytical lenses for various domains", "Reports findings alongside a report of analysis blind spots" ], "readmeHash": "cf47b29ff3a08b45468f975a5cdce30bfb939df82944ed736007719028b1f2f2", "generatedAt": "2026-07-27T18:52:53.951Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-29T14:18:15.448Z" }, "EfficientContext/ContextPilot": { "summary": "ContextPilot is a context optimization tool that sits between context assembly and inference to increase prefix reuse and eliminate duplicate blocks. It employs a context index to reorder and deduplicate text and multimodal data, aiming to reduce KV recomputation and token usage in long-context workloads.", "highlights": [ "Supports text and multimodal context reordering for prefix-cache sharing.", "Integrates with vLLM, SGLang, llama.cpp, OpenClaw, and Hermes Agent.", "Provides a context index for reordering, deduplication, and cache-aware scheduling." ], "readmeHash": "3d0f90f49ba5c3e887fb29c76a8813d4809fd8d02f19f1a661cd67c1f9e11cde", "generatedAt": "2026-09-25T05:08:30.510Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-28T16:08:28.112Z" }, "335234131/agent-browser-mcp": { "summary": "Agent-browser-mcp is a Model Context Protocol server that enables AI agents to interact directly with a user's active Chrome browser session rather than a sandboxed environment. It works through a local bridge and a Chrome extension to maintain existing login states, cookies, and open tabs while providing low-level control. The project allows agents to perform page scanning, execute JavaScript, and issue Chrome DevTools Protocol (CDP) commands. It also supports physical hardware interaction, including real mouse movements, keyboard inputs, and desktop screenshots.", "highlights": [ "Operates real Chrome sessions with preserved login states and cookies", "Executes native CDP commands for advanced DOM and browser control", "Supports physical mouse and keyboard inputs for human-like automation" ], "readmeHash": "9b786d4aecfbf43536158c1694a2f6770479e2412a1e405ef2c6471884da59c4", "generatedAt": "2026-05-01T16:47:44.339Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:07:07.347Z" }, "jwangkun/hermes-agent-guide": { "summary": "This project is a comprehensive Chinese-language guide designed to provide a systematic learning path for the Hermes Agent framework. It organizes fragmented documentation into 16 volumes covering the system's five-layer architecture, memory systems, and skill ecosystems. The guide includes practical tutorials for installation, MCP protocol integration, and multi-agent orchestration. It also provides specific migration paths for OpenClaw users and strategies for commercial application.", "highlights": [ "Detailed analysis of five-layer architecture and three-layer memory system", "Guides for 47 built-in tools and 6,000+ MCP service integrations", "Deployment tutorials for local, Docker, and VPS environments" ], "readmeHash": "7aefe614cf495f7eccee237a57731ffe6443ede45cdf6c1dbb37a9f05664bac1", "generatedAt": "2026-06-23T07:52:12.755Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:07:10.364Z" }, "longyunfeigu/learn-hermes-agent": { "summary": "Learn Hermes Agent is a code-first tutorial designed to teach the construction of production-grade autonomous AI agents using Python. It employs a 27-chapter structured path that guides users from building a basic conversation loop to implementing complex self-evolution mechanisms. The curriculum covers critical infrastructure including session persistence via SQLite, a multi-platform gateway for various messaging apps, and tool registry systems. It focuses on the architectural backbone required for agents to operate autonomously across different environments and persist knowledge over time.", "highlights": [ "Multi-platform gateway for Telegram, Discord, Slack, and WeChat", "Session persistence using SQLite with FTS5 for conversation memory", "Advanced capabilities including MCP integration and RL-based self-evolution" ], "readmeHash": "9b9cc1d587c7dca0a2b828c2efd1acb8b774355c8004fdd3a614f1155458a75b", "generatedAt": "2026-06-29T09:50:35.260Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:07:13.391Z" }, "jpalmae/hermeshq": { "summary": "HermesHQ is a Docker-based control plane designed to orchestrate and manage multiple Hermes Agent instances through a centralized web application. It provides an operational layer for the Hermes runtime, featuring task dispatch, multi-agent delegation, and administrative controls such as RBAC and permission policies.", "highlights": [ "Orchestrates multiple managed agents with separate workspaces and identities.", "Provides per-agent Telegram and WhatsApp communication channels.", "Enforces granular tool, path, and network permission policies." ], "readmeHash": "51ba4b43c5036f262a5759954816697f5239ddd7f505c941df159b9c065ac8e5", "generatedAt": "2026-09-24T17:21:54.943Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-28T16:08:28.819Z" }, "plastic-labs/honcho": { "summary": "Honcho is memory infrastructure for stateful agents, with a FastAPI core service and Python and TypeScript SDKs. It stores messages and events, processes them in the background to update peer representations, and provides APIs for querying context, search results, and natural-language answers. It can be used as a managed service, run locally, or self-hosted.", "highlights": [ "Tracks peers, sessions, and messages over time", "Provides Python and TypeScript SDKs", "Supports managed, local, and self-hosted use" ], "readmeHash": "83dd535c6825e3227fb21fac379665ef8476c6ca5d628fcd061f96d94b220afb", "generatedAt": "2026-10-23T17:57:35.780Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "adnw-vinc/hermes-nextcloud": { "summary": "This project is a Hermes Agent skill that enables direct interaction with self-hosted Nextcloud instances for personal data management. It interfaces with Nextcloud using WebDAV, CalDAV, CardDAV, and the Notes API to perform operations without requiring a browser or third-party sync services. Users can manage files, edit notes, schedule calendar events, track tasks, and search contacts through a command-line interface. The skill prioritizes security by utilizing Nextcloud App Passwords and storing credentials in a restricted local environment file.", "highlights": [ "Manages files, notes, calendars, tasks, and contacts via standard protocols.", "Communicates securely over HTTPS using Nextcloud App Passwords.", "Includes a guided setup script for credential validation and configuration." ], "readmeHash": "0508c60ebb9523b0cba6806d15b1e1c80244a481419332c8c1c916c18ac57967", "generatedAt": "2026-05-01T16:48:02.826Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:07:23.043Z" }, "Xquik-dev/hermes-tweet": { "summary": "Hermes Tweet adds Twitter search, reads, monitoring, and X actions to Hermes Agent through Xquik. It includes separate tools for reads and actions, with actions disabled by default, plus a bundled Hermes Skill and two slash commands.", "highlights": [ "Search X posts and read profile timelines.", "Actions include extraction, monitoring, posting, and replying.", "Slash commands show account status and current X trends." ], "readmeHash": "ca2a0c41d3b0a249a2001614fdc300cec73a4987deeb506c601c627c9e1b9c72", "generatedAt": "2026-10-19T15:06:48.675Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Context4GPTs/klodi-plugin": { "summary": "klodi is an Agent2Agent marketplace designed for AI agents to list, haggle, and close deals. It provides adapters for various agent hosts to enable agent-led negotiations and transactions.", "highlights": [ "Supports multiple host adapters including OpenClaw, Hermes, and Moltis", "Uses on-disk strategies to manage identity and ratings", "Facilitates a lifecycle of listings, offers, channels, and transactions" ], "readmeHash": "6b9d1812938047d57c8ba1630fd76ab4aee7672124b7420acc57e13a8450e5f8", "generatedAt": "2026-07-14T21:31:42.186Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:37:24.243Z" }, "yepyhun/Brainstack": { "summary": "Brainstack is a local-first memory kernel for Hermes agents that replaces flat memory storage with a composite provider model. It organizes information into distinct shelves for profiles, session continuity, temporal graph truth, and corpus retrieval to prevent prompt flooding and maintain factual consistency.", "highlights": [ "Integrates Hindsight, Graphiti, and MemPalace into a single memory owner.", "Uses SQLite, Kuzu, and Chroma for state and retrieval management.", "Provides a purpose-built evidence packet to the agent per query." ], "readmeHash": "706f629d18d3a11c13c6d26732d7fea947e755252052bf44dcc2dcd993d7495e", "generatedAt": "2026-07-14T21:31:35.177Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:37:24.212Z" }, "kfa-ai/hermes-timetree-sync": { "summary": "hermes-timetree-sync is a Python-based CLI and library that enables Hermes Agents to manage TimeTree calendar events without manual user intervention. Since TimeTree discontinued its official API, this bridge uses authenticated web endpoints and session cookies to perform non-interactive writes. It supports creating all-day events directly via HTTP requests and timed events through a Safari-backed automation path. The tool allows agents to translate natural language requests into structured calendar updates using locally stored credentials and optional label mapping policies.", "highlights": [ "Syncs all-day and timed events using TimeTree web endpoints", "Supports non-interactive batch writes for multiple calendar entries", "Maps natural language terms to calendar labels via YAML policy" ], "readmeHash": "66d2bd57bf6d8ed723705ddee94a59f1b066e9b6edf7390dd76921e2c16810c3", "generatedAt": "2026-05-25T09:24:34.940Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:07:36.876Z" }, "jau123/MeiGen-AI-Design-MCP": { "summary": "MeiGen AI Design MCP provides tools for AI assistants to search for visual inspiration and create images and videos. It also supports tasks such as background removal, image upscaling, and product imagery, with connections available through remote OAuth or a legacy local npm setup.", "highlights": [ "Search inspiration and view associated prompts.", "Generate images and videos with available models.", "Connect through remote OAuth or local npm." ], "readmeHash": "411a8a5947df3e1c087f0bc1a64611285eb488a19da2b984ada9ce663f0acd37", "generatedAt": "2026-10-21T10:07:23.212Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "aliasocracy/hermes-tool-slimmer": { "summary": "Hermes Tool Slimmer is a plugin for Hermes Agent designed to reduce the token overhead of model requests by trimming tool schemas. It uses BM25 ranking to select relevant tools based on the user's request while maintaining a fallback mechanism for the model to request the full tool catalog.", "highlights": [ "Reduces tool schema tokens using local keyword-based ranking.", "Integrates via middleware without patching Hermes core on version 0.19+.", "Includes a dashboard for monitoring savings and managing tool configurations." ], "readmeHash": "2d8f908e4b54d8d3ac535f0ebf28ec0c7b46668c594566e24210d1de18eb45a4", "generatedAt": "2026-10-01T06:11:06.797Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:04:16.660Z" }, "BORT-AGENTS/hermes-bort": { "summary": "hermes-bort is a plugin for Hermes Agent that enables interaction with BORT (BAP-578) agent NFTs on the BSC mainnet. It allows Hermes processes to read on-chain agent states, execute actions via operator-signed writes, and manage portable session memory anchored to IPFS. The plugin integrates with the hermes-agent-self-evolution optimizer to commit verifiable on-chain learning records and evolved skills directly to an agent's knowledge registry.", "highlights": [ "Reads full agent state, IPFS identity, and knowledge registries", "Executes on-chain actions via VaultPermissionManagerV2 with local policy gating", "Anchors local session memory to IPFS and KnowledgeRegistryV2" ], "readmeHash": "3c1ed1120ac681d518d474288c52711da4e44a068b498afe2adf14d97f37830e", "generatedAt": "2026-05-19T07:47:26.358Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:07:45.585Z" }, "volcengine/OpenViking": { "summary": "OpenViking is an open-source context database that organizes an AI agent’s knowledge, memories, and skills in a browsable virtual filesystem using `viking://` URIs. It provides directory summaries and three context layers—abstract, overview, and full details—so agents can assess relevance before loading source content.", "highlights": [ "Organizes resources, memories, and skills as files.", "Supports scoped semantic search across directories.", "Archives sessions and extracts inspectable Markdown memories." ], "readmeHash": "0887d2b5318c729363da726563ad69c68f1f852117d17aae5c85457799a814c9", "generatedAt": "2026-10-20T11:51:21.298Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "supermemoryai/supermemory": { "summary": "Supermemory describes itself as a memory and context layer for AI, with an API for adding memory, RAG, user profiles, connectors, and file processing. Its plugins and MCP server provide persistent memory for supported AI assistants, including Hermes.", "highlights": [ "Extracts conversation facts and handles contradictions and automatic forgetting.", "Connectors include Google Drive, Gmail, Notion, OneDrive, and GitHub.", "Processes PDFs, images, videos, and code." ], "readmeHash": "1b3e6d9804718bc94c6309e29a870ce6eabe3fedabb2739152ab85910a29b5f9", "generatedAt": "2026-10-19T15:06:49.813Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "campfirein/byterover-cli": { "summary": "ByteRover CLI is an interactive REPL and context management tool that provides AI coding agents with persistent memory and a structured context tree. It features a Git-like version control system for managing knowledge through commits, branches, and merges, while syncing data with a cloud platform for team collaboration.", "highlights": [ "Interactive TUI and REPL interface powered by React/Ink", "Git-like version control for context tree management", "Supports 20 LLM providers and 24 built-in agent tools" ], "readmeHash": "57ca305e6681f817d45b87db7ecacb1cdd9841b748782e3ff162624585343d17", "generatedAt": "2026-07-14T21:31:37.684Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:37:29.864Z" }, "yantrikos/yantrikdb-hermes-plugin": { "summary": "YantrikDB Hermes Plugin provides YantrikDB as a memory provider for Hermes Agent. Its embedded backend runs without a separate server, and the plugin supports memory recall with scoring reasons, duplicate canonicalization, and contradiction tracking.", "highlights": [ "Embedded backend is the default; an HTTP backend is also available.", "Recall results include a why_retrieved reason.", "Includes contradiction tracking and memory canonicalization." ], "readmeHash": "ab9a813f9b9c10ddf9121527a4f403026a59ffef9ee8046bad51c37b27b55619", "generatedAt": "2026-10-15T04:55:41.234Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:40.025Z" }, "Ladybug-Memory/hermes-memory-plugin": { "summary": "Hermes Ladybug Memory is a local-first memory provider plugin that enables Hermes agents to store and retrieve information without relying on cloud services or external APIs. It utilizes a columnar embedded graph database to manage typed memory entries, named relationships, and importance-weighted recall. The system prioritizes high-value information using a 1-10 importance scale and supports BM25 keyword search for precise retrieval. Users can also enable optional entity extraction via GLiNER2 to build structured knowledge graphs directly on their local machine.", "highlights": [ "Local-only storage using a single .lbdb embedded graph database file", "Importance-weighted recall surfaces high-priority memories during context prefetch", "Supports BM25 keyword search and named graph relationships between entries" ], "readmeHash": "c256453924e81874712dd719696ddf617e9c3cf7aba60d762124cd3ec38f8b0f", "generatedAt": "2026-05-24T11:48:52.880Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:08:02.570Z" }, "MukundaKatta/hermes-agentmemory": { "summary": "This is a standalone memory plugin for Hermes Agent built on agentmemory. It uses a pull-model for episodic memory that performs synchronous writes and immediate deletions without leaving derived summaries behind.", "highlights": [ "Provides an audit trace of prefetch operations via trace.jsonl", "Supports real, immediate event deletion", "Uses an on-demand summarizer" ], "readmeHash": "0f3f86d012e00acd594f7d92e660cf89b3f7b6e56dd483d29490451cf03c930b", "generatedAt": "2026-07-14T21:31:59.084Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:37:29.722Z" }, "Signet-AI/signetai": { "summary": "Signet stores and syncs memories, prompts, transcripts, institutional knowledge, and secrets across AI tools and models. It imports context from connected sources and files, then maintains a structured map with citations linking information to its original evidence.", "highlights": [ "Connects to agent harnesses through hooks, plugins, or extensions.", "Provides a dashboard for browsing memory and asking questions.", "Can run on a local machine or as a team server." ], "readmeHash": "6e61a31c5d4807d2b3761b035b51c46e60f335e0271b22fee72c95fe3ab02d71", "generatedAt": "2026-10-19T18:21:47.200Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "linke-ai/hermes-agent-team": { "summary": "Hermes Agents Team is a local web-based system designed to orchestrate multi-agent collaboration using Hermes Agent profiles. It utilizes a Leader/Specialist architecture to automate task decomposition, assignment, and review through a centralized Web UI and Hermes Kanban. The system integrates MCP and ACP protocols to manage agent communications, skills, and tool calls entirely within a local environment. Users can monitor real-time dialogue, terminal outputs, and task progression while maintaining full control over model configurations and data privacy.", "highlights": [ "Orchestrates Leader and Specialist agents for automated task lifecycle management", "Integrates MCP and ACP protocols for local tool and communication handling", "Real-time Web UI for monitoring agent dialogues and Kanban task flows" ], "readmeHash": "9ca8e430f0d7517d3a339c08fb0a8f77ed9bb5e9041957220d5c8d9b6de3f1d8", "generatedAt": "2026-05-28T12:12:26.139Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:08:11.769Z" }, "Shelpuk-AI-Technology-Consulting/kindly-web-search-mcp-server": { "summary": "Kindly Web Search is an MCP server designed to provide AI coding agents with structured content and documentation from the web. It retrieves full conversations, code snippets, and metadata from sources like StackExchange and GitHub Issues in a single call, rather than returning only search snippets.", "highlights": [ "Integrates with StackExchange, GitHub, arXiv, and Wikipedia APIs.", "Uses a headless browser for real-time webpage parsing and extraction.", "Supports multiple search providers including Serper, Tavily, and SearXNG." ], "readmeHash": "385cbc40b83fc297cd777193b815fa37cb3328546fd1b289b3ae2b90d33fa1b0", "generatedAt": "2026-09-28T16:11:03.573Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:04:23.730Z" }, "colbymchenry/codegraph": { "summary": "CodeGraph is a local code graph for coding agents, with a CLI that builds a project graph in a `.codegraph/` directory. It can connect to supported agents through an MCP server and automatically update the graph when project files change.", "highlights": [ "Builds a project graph with `codegraph init`.", "Auto-syncs the graph when files change.", "Connects to agents through an MCP server." ], "readmeHash": "a0fedddce63b47d82be01bb3470ba01f610ee34d16a55bb393663b14d489a908", "generatedAt": "2026-10-19T15:06:28.867Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Felix-Forever/hermes-agent-desktop": { "summary": "Hermes Agent Desktop is a native GUI client that transforms the Hermes Agent from a single-agent CLI into a collaborative multi-agent system. It utilizes a Project Manager orchestrator to decompose complex requirements and delegate tasks across a team of 20 specialized AI agents. The application is built with pywebview and supports any OpenAI-compatible LLM provider. Users can manage agents via a CRUD dashboard and install capabilities through an integrated visual Skill Store.", "highlights": [ "Orchestrated collaboration between 20 specialized AI agents", "Visual Skill Store with 50+ curated one-click installations", "Native desktop interface with streaming SSE responses and model switching" ], "readmeHash": "70d5f367805b9e2ab7d98574fa9ad1ea283ecdfba82414f082a9558fe332e0ca", "generatedAt": "2026-06-01T09:55:23.748Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:08:32.425Z" }, "AkoliteZA/hermes-agent-idea-workflow": { "summary": "The Hermes Agent Idea Workflow is a suite of skills designed to transform vague product concepts into structured, build-ready specifications. It operates through a staged pipeline that separates product intent from technical implementation, utilizing guided interviews and a reusable question bank to generate markdown artifacts. The system supports both a Lite mode for quick capture and a Full mode that produces comprehensive design docs, UI briefs, and implementation specs. These outputs are specifically formatted to serve as direct handoff material for downstream coding workflows like Superpowers for GPT.", "highlights": [ "Converts rough ideas into structured design and implementation documents", "Features Lite and Full modes for varying project complexity", "Includes a spec review gate with PASS/FAIL evaluation models" ], "readmeHash": "81dd895318c6a3514436b49f7e5bb0d31a2f4cef5a440bffdec5a1b55b4a5f01", "generatedAt": "2026-05-28T12:12:52.995Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:08:35.265Z" }, "shannhk/hermes-agent-control-room": { "summary": "Hermes Agent Control Room is a template-based framework designed to organize and govern Hermes agents across VPS environments. It functions as a sidecar control plane that provides a structured system map, registry, and runbook library to manage agent operations. The architecture supports scaling from a single agent to specialized teams using a task bus pattern for delegation and synthesis. Users can interact with agents directly or through an optional orchestrator that serves as a central front door for automated workflows.", "highlights": [ "Level-based architecture for scaling from single agents to automated teams", "Task bus pattern for orchestrator-to-specialist delegation and handoffs", "Bundled skills for VPS provisioning, security auditing, and backup management" ], "readmeHash": "72d8e831adf5636ea04937e563a2ba4ee7814941cc93a21f4da495729830e648", "generatedAt": "2026-05-28T12:12:56.652Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:08:38.115Z" }, "Salomondiei08/oh-my-hermes": { "summary": "Oh My Hermes is a workflow layer for Hermes Agent, installed with a shell script. It provides skills and workflows for product discovery, design, building, review, deployment, and ongoing operations, with a CTO role that delegates to seven focused agents through Hermes Kanban.", "highlights": [ "Installable with install.sh.", "Includes skills for product development and operations.", "Supports a CTO loop with seven focused agents." ], "readmeHash": "24854726d1038e613090c7c5ebfec43a39a016f736cdb772bce7e385dbaa26a6", "generatedAt": "2026-10-19T15:06:28.767Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "liaohch3/claude-tap": { "summary": "claude-tap is a local proxy and trace viewer for inspecting API traffic and agent context from AI coding clients. It can run a client through the proxy or listen to local app transcripts, and can export trace sessions as self-contained HTML viewers.", "highlights": [ "Inspect prompts, tool calls, streaming responses, and token usage.", "Compare adjacent requests to see changes in context.", "Redacts common auth headers before recording traces." ], "readmeHash": "2f068ba80664a6604dd84c906bfd7294a45a62022a6929248eb94e7367ba696a", "generatedAt": "2026-10-17T15:17:15.571Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:44.353Z" }, "jazzyalex/agent-sessions": { "summary": "Agent Sessions is a macOS app for searching local histories from 16 coding agents, reading transcripts, and finding supported image references. It can provide resume commands or open supported CLI sessions in terminal apps, and tracks per-session Codex and Claude quota use; session history and the search index stay on the Mac.", "highlights": [ "Search prompts, tool calls, errors, and file paths across local histories", "View session metadata and supported token or pricing details", "No session history uploads" ], "readmeHash": "025000f068f405b03e583fb963907810f48b0104488f0e8a34d52006515c7da5", "generatedAt": "2026-10-25T05:17:36.473Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "vectrocomputers/Hermes-Agent-Action-Plan": { "summary": "The Hermes Agent Action Plan is a configuration framework and operational guide designed to optimize Hermes Agent deployments on Ubuntu 24. It provides a structured philosophy for setting up an autonomous assistant that prioritizes local-first tools, persistent memory, and granular system access. By implementing specific security hardening, firewall rules, and communication guidelines, the plan enables the agent to perform complex tasks while maintaining user privacy and auditability. The guide serves as a template for transitioning the agent from a basic chatbot into a capable system administrator or personal assistant.", "highlights": [ "Privacy-first configuration prioritizing local and open-source tools", "Structured memory system using daily logs and long-term curated files", "Security-focused system access strategy with audit trails and recovery steps" ], "readmeHash": "7e0b349d62a0153026d74c6a94ebf2d41a5cbd6f39b99363f6f7f2b0efe6eb51", "generatedAt": "2026-05-28T12:13:11.858Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:08:52.739Z" }, "xiaonancs/hermes-agent-study": { "summary": "This project provides a comprehensive technical analysis and documentation series focused on the internal architecture of the Hermes Agent by Nous Research. It breaks down the system's core components, including its Python-based execution loop, tool registry, and self-evolution engine, while providing nearly 100 architectural diagrams. The research also explores the project's lineage by comparing its structure to OpenClaw and investigating claims regarding its EvoMap-inspired self-improvement logic. Additionally, it tracks rapid version iterations from v0.9.0 to v0.14.0, covering new features like the Transport ABC and autonomous curators.", "highlights": [ "Deep source-code analysis of Hermes Agent architecture and evolution engine", "Comparative studies against OpenClaw, EvoMap, and other Agent frameworks", "Detailed documentation of memory systems, skill hubs, and tool registries" ], "readmeHash": "b8b65a338954ce2376cb33bd4927d01f52cb64d11efbe102a1684f321ba91b75", "generatedAt": "2026-05-28T12:13:15.883Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:08:55.689Z" }, "carterwayneskhizeine/hermes-agent-windows-R": { "summary": "This project is a Windows-native fork of Hermes Agent designed to resolve compatibility issues with path handling, process management, and terminal encoding. It implements specific adaptations for Git Bash path conversion and introduces a multi-profile runtime system for running parallel instances. The fork includes a dedicated PowerShell-based startup workflow and a web-based Dashboard UI optimized for the Windows environment.", "highlights": [ "Fixes Git Bash path handling and terminal encoding for Windows", "Supports multi-profile management and parallel gateway instances", "Includes psmux integration for one-click automated environment launching" ], "readmeHash": "8021943ded8b50b2bfaaccb6d99d45569cb72634e04604b2e3b38d56256331db", "generatedAt": "2026-05-28T12:13:19.637Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:08:58.643Z" }, "sheawinkler/hermes-agent-ultra": { "summary": "Hermes Agent Ultra is a Rust-based autonomous agent runtime designed to provide feature parity with the original Hermes Agent while adding enhanced reliability and operator control. It achieves this through a fully Rust-native core that implements deterministic execution paths and a dedicated safety policy engine. The project offers advanced session management features like time-travel debugging and provides extensive support for local and self-hosted inference backends. It is optimized for production operations through deep diagnostics, observability tools, and automated parity synchronization.", "highlights": [ "Rust-native runtime with deterministic execution paths", "Extensive support for local and self-hosted inference backends", "Advanced session branching and time-travel debugging" ], "readmeHash": "42139226c83f9cc6b84d5f09954a2d33b4d313fdb28f2b276896b7cc7eea2af2", "generatedAt": "2026-07-14T20:16:00.830Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:28:47.595Z" }, "AbuZar-Ansarii/Hermes-Agent-On-Android": { "summary": "This project provides a Termux-based installer and guide to deploy the Nous Research Hermes Agent on Android devices. It includes a one-line installation script that sets up a Linux environment via proot-distro to run the AI framework's interactive interface and gateway service.", "highlights": [ "Automated installation script for Termux using proot-distro Ubuntu.", "Supports interactive model setup, gateway services, and shell command execution.", "Compatible with external providers and local models via Ollama." ], "readmeHash": "64c9decf9227370f8976f596d1186d69ad01c225d1e1d473f2f1937a0142e672", "generatedAt": "2026-09-14T14:31:16.909Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:36:33.588Z" }, "jefferyjob/awesome-hermes-agent-zh": { "summary": "This repository is a Chinese translation of the awesome-hermes-agent list, providing a curated collection of skills, tools, and resources for the Hermes Agent ecosystem. It includes links to official Nous Research resources, community-developed skills, and various integrations.", "highlights": [ "Curated list of Hermes Agent skills and tools", "Includes official Nous Research resources", "Categorized by maturity levels" ], "readmeHash": "08b83080dc9d953c931ff6ac27e267d60958adde2c80690becb1fe63b8c47879", "generatedAt": "2026-08-04T18:54:47.476Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-10T08:29:31.310Z" }, "anneheartrecord/hermes-agent-anatomy": { "summary": "This project provides a systematic technical decomposition and architectural analysis of the Hermes Agent source code. It breaks down the framework's 334,000 lines of Python code into modular guides covering the core agent loop, tool registration, and multi-provider adapters. The documentation includes visual technical illustrations and a 13-dimension comparative study against other frameworks like Claude Code and OpenClaw. Users can leverage these docs to understand complex features such as context compression, RL training data collection, and 11-platform message gateway integration.", "highlights": [ "Detailed analysis of core agent cycles and iteration budget management", "Deep dives into context compression and structured LLM summarization techniques", "Technical breakdown of 8 memory backends and 11 message platform adapters" ], "readmeHash": "e30062f98cdc333b42a2ba4ef2be353b0a343d92bd3e164f991743efac79eaaa", "generatedAt": "2026-05-28T12:13:34.607Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:09:15.064Z" }, "nickvasilescu/hermes-desktop-os1": { "summary": "Hermes Desktop - OS1 Edition is a native macOS workspace designed to manage Hermes Agents running on Orgo cloud computers or SSH hosts. It provides a unified interface for VM provisioning, one-click agent installation, and real-time terminal access via websockets. The application includes built-in tools for session management, Kanban boards, and file editing with conflict checks. Additionally, it features a WebRTC-based voice mode that integrates OpenAI's Realtime API with Orgo MCP tools for voice-controlled agent interactions.", "highlights": [ "Native macOS interface for Orgo cloud VMs and SSH hosts", "One-click automated Hermes Agent installation on fresh cloud computers", "Integrated WebRTC voice mode using OpenAI Realtime and Orgo MCP" ], "readmeHash": "188e6bf775da129fa40a4f6921aeaba3563ea2423c7f2a6c3c673e5678eebd5b", "generatedAt": "2026-05-28T12:13:38.422Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:09:17.718Z" }, "adityahimaone/hermes-agent-rtk-caveman": { "summary": "This project provides a preconfigured setup for Hermes Agent that integrates Rust Token Killer (RTK) and Caveman templating to reduce token consumption during CLI operations. It works by wrapping standard terminal commands and outputting compact data structures, achieving a reported 90-99% reduction in token usage. The repository includes over 40 pre-configured skills covering finance management, GitHub workflows, research tools, and DevOps automation. Users can deploy the environment via an automated installer to gain immediate access to token-optimized aliases for git, linting, and testing tasks.", "highlights": [ "Achieves 90-99% token savings using RTK and Caveman templating", "Includes 40+ pre-configured skills for finance, devops, and research", "Automated cross-platform installation for macOS and Linux systems" ], "readmeHash": "8023cc4f82acf65357f4a8649a38ec35bc7fd88e4954a5ac44be6cfd42600dac", "generatedAt": "2026-05-28T12:13:42.472Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:09:20.489Z" }, "pengchengxia75-arch/hermes-agent-windows": { "summary": "This project is a Windows-native adaptation of the Hermes Agent that eliminates the need for WSL2 or Linux environments. It utilizes a PowerShell-based installer to deploy the agent directly into the Windows ecosystem while resolving platform-specific issues like process signal handling. The fork introduces a dedicated Web UI management panel for real-time chat, session history, and visual configuration of skills and API keys.", "highlights": [ "Native Windows installation via a one-line PowerShell script", "Eliminates WSL2 requirement for running Hermes Agent", "Integrated Web UI for chat, logs, and token usage analytics" ], "readmeHash": "24bbab24262c6e1d45e174248aecb625be1dbd107eb38e242ab6e22719d5b7b8", "generatedAt": "2026-05-28T12:13:45.926Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:09:23.438Z" }, "ultraworkers/hermes-agent-helm-chart": { "summary": "This unofficial community Helm chart facilitates the deployment of Nous Research's Hermes Agent on Kubernetes clusters. It provides a structured configuration model that enforces state-safety guardrails, such as single-replica constraints and specific update strategies, to prevent data corruption in persistent environments. The chart supports both direct workload management and an operator-ready mode for platform teams using custom resource definitions. It is optimized for multi-tenant isolation, offering granular control over networking, storage, and secret management for individual agent instances.", "highlights": [ "Enforces single-writer state safety for persistent HERMES_HOME storage", "Supports direct deployment or operator-ready CRD-based workflows", "Configurable gateway mode with optional API and webhook listeners" ], "readmeHash": "b5206b4ad19a5e4fee74f3915521309ed746d75f57119784ab400cc44de5d50d", "generatedAt": "2026-05-28T12:13:49.699Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:09:26.427Z" }, "ZeroPointRepo/youtube-skills": { "summary": "YouTube Skills provides a set of tools for AI agents to access YouTube transcripts, search videos, browse channel data, and extract playlist contents. It operates via the TranscriptAPI backend, requiring an API key but functioning without yt-dlp, headless browsers, or binaries.", "highlights": [ "Supports transcripts, video search, channel browsing, and playlist extraction.", "Compatible with Hermes Agent, OpenClaw, Claude Code, and Cursor.", "Requires a TranscriptAPI key with a 100-credit free tier." ], "readmeHash": "0cdaefaf0e7929871eff5df5ea622d027940a61e592f52a0fe0a60c519e7ec04", "generatedAt": "2026-08-23T07:08:27.424Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-24T08:33:15.187Z" }, "stainlu/hermes-labyrinth": { "summary": "Hermes Labyrinth is a read-only observability plugin for Hermes Agent that visualizes autonomous workflows as a structured map of events. It functions as a black-box recorder, reading local state to track prompts, tool calls, model transitions, and subagent activities without mutating session data. The tool provides deep diagnostic visibility through specialized views for skill overrides, scheduled cron tasks, and redacted journey reports. Users can export evidence in Markdown or JSON formats while relying on built-in redaction to protect sensitive credentials.", "highlights": [ "Visualizes agent journeys through ordered crossings and event maps", "Provides diagnostic views for skill overrides and cron schedules", "Generates redacted Markdown and JSON reports for session evidence" ], "readmeHash": "1b12e95cd3cac4e0c493b5c96af51f34cd5edd1bcb4c316832dff19a3c2106f3", "generatedAt": "2026-05-28T12:13:56.920Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:09:32.664Z" }, "marshallrichards/ClawPhone": { "summary": "ClawPhone provides scripts and configuration guides for deploying agent CLIs on Android devices. It leverages Termux to run these agents natively, utilizing proot or compatibility layers only when necessary. The project includes a streamlined installer for Google's Antigravity CLI and detailed implementation notes for running OpenClaw. This setup allows agents to operate in the background with potential access to phone hardware via Termux:API and Termux:GUI.", "highlights": [ "Automated installer for Antigravity CLI on ARM64 Android", "Configuration guides for running OpenClaw in Termux", "Hardware and UI access via Termux:API and Termux:GUI" ], "readmeHash": "c79df82caafcb116bf8961ae001a7bfba82e2a6621b6ae7c0afec5e35089cc82", "generatedAt": "2026-06-08T09:41:45.255Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:09:35.157Z" }, "agent-of-empires/agent-of-empires": { "summary": "Agent of Empires is a session manager for AI coding agents on Linux and macOS. It provides TUI, web, CLI, and HTTP API surfaces for running agents in persistent tmux sessions, with options including Git worktrees and container isolation.", "highlights": [ "Persistent tmux sessions with agent status detection and notifications", "Supports Git worktrees and multi-repository workspaces", "Web dashboard includes structured views, terminal access, and diffs" ], "readmeHash": "cc6571818fe2a78a2f9e273bc210c18337bab6667cd9ce0cdbbcbff7a9505125", "generatedAt": "2026-10-15T17:12:47.167Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:50.590Z" }, "nexu-io/open-design": { "summary": "OpenDesign is a collaborative AI design workspace where teams can start with a brief, create a prototype that follows a design system, share it for feedback, and refine it with an agent. It also supports creating and refining slides, images, and motion graphics, and provides skills, design templates, design systems, and plugins under the Apache 2.0 license.", "highlights": [ "Create and refine interactive web prototypes with a live preview.", "Create slide decks, images, and motion graphics.", "Includes skills, design templates, design systems, and plugins." ], "readmeHash": "97c4ced112de9039a62369d78bf6cbdd6f50e6ed77798aa44bba87a746417156", "generatedAt": "2026-10-19T15:06:34.927Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "farion1231/cc-switch": { "summary": "CC Switch manages API providers for Claude Code, Claude Desktop, Codex, Gemini CLI, Grok Build, OpenCode, OpenClaw, Hermes Agent, Pi, MiniMax Code, Kimi Code, and Oh My Pi. It also provides one place to manage MCP, Skills, and Prompts, without hand-editing JSON, TOML, or YAML configuration files.", "highlights": [ "Switch API providers in one click.", "Manage MCP, Skills, and Prompts in one place.", "Available for Windows, macOS, and Linux." ], "readmeHash": "4be9178443555a92cc31cb2deba54ea562b5229ab70aec79873b1646363bfe7c", "generatedAt": "2026-10-24T11:19:02.996Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "iOfficeAI/AionUi": { "summary": "AionUi is an open-source cowork app where AI agents can work with files, write code, browse the web, and automate tasks with user approval. It includes a built-in agent with file access, web search, image generation, and MCP tools, and offers a unified interface for using external agents.", "highlights": [ "Includes 21 built-in professional assistants.", "Supports remote access through WebUI and messaging platforms.", "Runs on macOS, Windows, and Linux." ], "readmeHash": "c72cd625c8e1ed4d23956d5320daac2824604e7c3d69b46eeaed6e7a7a8d8954", "generatedAt": "2026-10-23T15:08:30.835Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "futurebrowser/hermes-exabase-plugin": { "summary": "This plugin integrates the Exabase M-1 memory engine into Hermes Agent to provide persistent, long-term context for AI interactions. It functions by storing conversation turns and inferred memories in a self-organizing knowledge graph that resolves contradictions and evolves over time. Users can configure retrieval precision, enable query expansion, and utilize result reranking to fine-tune how the agent accesses stored facts and preferences. The integration includes dedicated tools for manual memory storage and searching, while also automatically processing conversation history in the background.", "highlights": [ "Integrates Exabase M-1 for self-organizing long-term agent memory.", "Provides tools for searching and manually storing specific memories.", "Supports configurable query expansion and result reranking for retrieval." ], "readmeHash": "c1d8b945396c8b0f447b710d6990d55fe37b717dd583615a180a3f39896eacde", "generatedAt": "2026-06-03T12:24:01.269Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:09:52.502Z" }, "willingning-coder/eagle-eye": { "summary": "Eagle Eye is a plugin for Hermes Agent designed to optimize skill selection by filtering large skill libraries into a small set of relevant candidates. It utilizes a five-layer retrieval architecture that combines deterministic keyword triggers, FTS5 text similarity, synonym matching, and dense semantic embeddings via RRF fusion. This approach reduces token consumption and prevents LLM confusion by providing only the most relevant skill hints rather than the entire list.", "highlights": [ "Five-layer retrieval including hard triggers and semantic embeddings", "Reduces token usage by narrowing skill lists before LLM calls", "Graceful degradation if embedding models or dependencies are missing" ], "readmeHash": "c6fa0475180fc8dce5424acd5b5a0e29f3a6168df6f6c510060e829d528839ac", "generatedAt": "2026-06-29T09:52:08.824Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:09:55.578Z" }, "liftaris/herm": { "summary": "Herm is a terminal dashboard and chat interface for Hermes Agent, built with OpenTUI and Bun. It supports chatting and session management, inspecting Hermes home settings and operational surfaces, and managing kanban work without leaving the terminal.", "highlights": [ "Connects to the Hermes Agent gateway.", "Includes tabs for profiles, sessions, skills, cron, and memory.", "Provides kanban boards, task details, diagnostics, and dispatch controls." ], "readmeHash": "bf26c54ae57f9d083d526bf8192042ecf694f77cc95d30499bc4ccc14038f574", "generatedAt": "2026-10-24T04:50:29.684Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "basilisk-labs/agentplane-hermes-plugin": { "summary": "This plugin acts as a transport adapter to bridge AgentPlane with Hermes as a native worker-lane. It manages the execution of AgentPlane tasks while Hermes maintains ownership of Kanban task state and LLM episodes.", "highlights": [ "Implements a native Hermes command group for AgentPlane tasks.", "Provides a conversational bridge for task approvals within Hermes.", "Requires AgentPlane version 0.7.6 or newer." ], "readmeHash": "e24a98a3c770547771209b74f85924c003a345eb3930235eac2ee53c296d42f7", "generatedAt": "2026-08-18T07:27:35.977Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-24T08:33:20.867Z" }, "8bit64k/cronalytics": { "summary": "Cronalytics is a plugin for Hermes Agent that provides observability and cost attribution for scheduled cron jobs. It utilizes the on_session_end hook to capture usage data and store it in a local SQLite fact database. Users can monitor job health and expenses through a visual dashboard, a programmatic CLI, and a specialized agent skill for automated diagnostics. This allows developers to identify failure patterns and optimize the economics of their agentic automations.", "highlights": [ "Tracks session-level usage and estimated costs for cron-originated runs", "Provides visual insights via a dedicated Hermes dashboard tab", "Includes a diagnostic agent skill for anomaly detection and recommendations" ], "readmeHash": "1884c80eadb8af2dafb852b9b03fe4df463537fc556fb4b5561d9a0669dcd185", "generatedAt": "2026-06-23T07:53:12.128Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:10:04.262Z" }, "Sahil-SS9/MrHermagi-tutorbot": { "summary": "MrHermagi-tutorbot is a Discord AI tutor bot that delivers daily lessons to a forum channel at 07:00 UK, with each week getting one thread and lessons posted as replies. Lessons come as a scannable Discord summary, a dark-mode HTML lesson, and TTS audio, with curriculum defined in a YAML file that can be swapped to change subjects. It runs on Hermes Agent profiles with a cron job and requires a Discord bot with Send Messages, Create Threads, and Attach Files permissions.", "highlights": [ "Daily lessons delivered to a Discord forum at 07:00 UK", "Curriculum defined in YAML; swap subjects without infrastructure changes", "Includes HTML lessons, TTS audio, and a Q&A channel" ], "readmeHash": "4664b1eac1bff293d1e23742467efb561933dc90af20c3b62de2030ae49cb1f5", "generatedAt": "2026-10-12T00:26:52.667Z", "model": "deepseek/deepseek-v4.1-flash", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:04:35.492Z" }, "Sahil-SS9/hermes-Custom-CLI-Themes": { "summary": "Hermes Custom CLI Themes is a collection of visual skins designed to personalize the Hermes Agent terminal interface. Each theme is defined in a single YAML file that configures color palettes, spinners, branding, and ASCII banner art. Users can apply these skins by placing them in a specific local directory and updating their configuration or using an in-app command. The project includes a variety of themed presets ranging from tactical and sci-fi aesthetics to minimalist investigator styles.", "highlights": [ "Provides YAML-based skins for terminal palettes and ASCII branding.", "Includes a one-liner script for automated skin installation.", "Supports custom extended keys for advanced UI element styling." ], "readmeHash": "ba6b51a542529c917da8d64a4212f48db50a1a6eef188a2353dd73fd17fa0bfb", "generatedAt": "2026-06-03T12:24:35.192Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:10:12.583Z" }, "mag3nt-com/openclaw-skill": { "summary": "mag3nt-pay is an agent payment skill that enables AI agents to autonomously pay for APIs that return HTTP 402 Payment Required errors. It works by detecting payment protocols from response headers, decoding the payment challenge, and settling the transaction using USDC via Mag3nt virtual cards. The skill supports Base, Solana, and most EVM networks, allowing agents to handle micropayments and subscriptions without human intervention. It is compatible with OpenClaw, Hermes Agent, and any framework following the agentskills.io standard.", "highlights": [ "Automates payments for HTTP 402 responses using USDC", "Supports Pay Link, x402, MPP, and AP2 protocols", "Compatible with OpenClaw, Hermes Agent, and agentskills.io standard" ], "readmeHash": "2dddc71f85f03a2d167b935f8f5ea3e0a7b357e56d160edee557c0c736cb6d49", "generatedAt": "2026-06-07T20:07:44.883Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:10:15.217Z" }, "indranilbanerjee/digital-marketing-pro": { "summary": "Digital Marketing Pro is an open-source, MIT-licensed AI marketing plugin with skills and specialist agents for brand setup, campaign planning, SEO audits, competitor analysis, and pre-publish checks. It guides brands through a 12-part engagement flow and supports multiple agent platforms, including Claude Code, Anthropic Cowork, OpenAI Codex, Cursor 2.5+, and Google Antigravity 2.0.", "highlights": [ "Includes 164 skills and 24 specialist agents.", "Live writes require typed approval or an approved standing rule.", "Supports 9 native platforms and 35+ Agent Skills platforms." ], "readmeHash": "e9396ce95cc3c24747b3a4f648cdb4229bc52111c3b9a1be909499c63eb46d6b", "generatedAt": "2026-10-12T17:07:28.221Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:58.631Z" }, "indranilbanerjee/socialforge": { "summary": "SocialForge automates monthly social media calendars by parsing content, matching brand assets, creating visuals, adapting copy for platforms, and preparing review galleries and delivery documents. It supports AI image generation with Vertex AI, AI video generation with WaveSpeed, and C2PA signing before review; paid generation requires a price quote and user approval.", "highlights": [ "Adapts copy for Instagram, TikTok, LinkedIn, Threads, X, Facebook, and YouTube Shorts", "Keeps brand photos pixel-faithful during asset-first compositing", "Pauses for client review before packaging delivery" ], "readmeHash": "15012cedc3507e89b777e8a1d1cdfc4669bccd0d2bfbe727c4451533a5137e69", "generatedAt": "2026-10-12T17:07:28.478Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:57.904Z" }, "Sahil-SS9/hermaguard": { "summary": "HermaGuard v2.1.0 is a read-only adversarial bug-hunting tool for AI agents that uses a deterministic pre-scan followed by three parallel subagents to identify code vulnerabilities. It features a verification sandbox for proof-of-concept execution and a consolidator that triages findings into markdown and JSON reports.", "highlights": [ "Parallel subagents for edge cases, adversarial review, and integration analysis.", "Deterministic pre-scan using semgrep, bandit, eslint, gosec, and ruff.", "Verified findings via proof-of-concept execution in a hardened sandbox." ], "readmeHash": "1dd7d18bb8b845a8cef1d86971af239ad8e5507f150a47e0629ec70cd7c99bb2", "generatedAt": "2026-08-28T18:49:49.890Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-31T15:34:20.987Z" }, "Sahil-SS9/hermes-simplify-swarm": { "summary": "Simplify Swarm is a multi-agent skill for code simplification that uses three parallel, read-only agents to analyze git diffs. The system categorizes findings into SAFE, CAREFUL, and RISKY tiers and requires user approval before applying changes.", "highlights": [ "Uses Hygiene, Clarity, and Correctness agents", "Supports TypeScript, JavaScript, Python, Go, and Rust", "Features a tiered application process with an approval gate" ], "readmeHash": "34b81694d63c06591bc36b73c9033e73019f435ba210c55537b0332720f164e8", "generatedAt": "2026-08-17T02:00:33.965Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-17T08:28:53.811Z" }, "Sahil-SS9/hermes-memlock": { "summary": "MemLock is a plugin for the Hermes ecosystem designed to prevent instruction decay following context compaction. It detects compaction events, audits whether pinned instructions remain in the active context region, and rehydrates missing instructions as a reminder block.", "highlights": [ "Detects Hermes context compaction markers to identify instruction drift.", "Audits active context regions for pinned anchors using keyword probes.", "Rehydrates drifted instructions only when they are missing from active memory." ], "readmeHash": "063e329d57939b298acfe97e2726d594e3093780c1b981130f464afa110b674b", "generatedAt": "2026-08-24T19:07:08.638Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-31T15:34:23.155Z" }, "Sahil-SS9/Toolaria": { "summary": "Toolaria is a single-purpose Hermes Agent plugin that stores oversized MCP or web tool results in a SHA256-addressed blob store and returns a compact excerpt with a fetch handle. The model can retrieve selected slices using `rescuer_fetch`, inspect structural outlines, search content, or pass the full result to another tool by reference.", "highlights": [ "Rescues oversized tool results before they enter the context window.", "Supports outline, search, range, grep, stat, and full fetch modes.", "Passes stored results to downstream tools using `tla:`." ], "readmeHash": "f79f0860685104872d55bad789d3aa0bddb6cac4521c9cf657ebda1bfbf42aae", "generatedAt": "2026-10-14T05:06:32.134Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:03:59.572Z" }, "hlothaire/hermes-trace": { "summary": "Hermes Trace is a plugin that provides visibility into agent execution by building structured trace graphs of session activity. It utilizes 18 lifecycle hooks to automatically capture LLM calls, tool usage, subagent spawns, and approval requests. Users can analyze these traces via a dedicated CLI or an in-session slash command to identify bottlenecks and track token usage. The system ensures reliability by wrapping all hooks in error-handling blocks and requires no runtime dependencies beyond the Python standard library.", "highlights": [ "Captures 18 lifecycle hooks for comprehensive event tracing", "Exports traces as JSON, Mermaid diagrams, and text trees", "Includes ASCII Gantt timelines for analyzing span concurrency" ], "readmeHash": "d92cbd930c69573b8a4a6cbfb5da547f19a0d8c54a8ee12bd7db363652beaac4", "generatedAt": "2026-06-19T08:22:19.636Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:10:39.424Z" }, "nujovich/hermes-telemetry": { "summary": "hermes-telemetry is a Hermes Agent plugin that records API and tool-call usage, including tokens, estimated cost, and latency. It enforces budget limits by notifying at a soft limit and blocking further tool calls at a hard limit; a breached cron-job budget also pauses that job.", "highlights": [ "Records usage in SQLite with WAL mode.", "Provides `/stats`, `/budget`, and `/setup` surfaces.", "Automatically syncs OpenRouter pricing." ], "readmeHash": "a89df8a7d8aea45ead233e4db7cced87293af1bec89994459527f647b4f46a45", "generatedAt": "2026-10-22T17:23:16.414Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "MnrGreg/hermes-venice-web": { "summary": "This plugin registers Venice AI's /augment endpoints as native web_search and web_extract backends within Hermes. It allows agents to use Venice for web searching and scraping through standard Hermes web tool infrastructure.", "highlights": [ "Integrates Venice AI /augment/search and /augment/scrape endpoints", "Supports Brave or Google as search providers", "Compatible with any LLM provider" ], "readmeHash": "a3b7b934b3afdd0def53e1e3025ad064ca2c91b3cc91dfda94b7070101bbd1d6", "generatedAt": "2026-07-27T01:02:39.186Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-27T09:02:54.464Z" }, "yonro/hermes-xmemo-plugin": { "summary": "hermes-xmemo is a native XMemo memory provider for Hermes Agent. It supports pre-turn context recall, explicit memory tools, mirroring Hermes built-in memory writes to XMemo, and session-end snapshots.", "highlights": [ "Runs inside Hermes and communicates with XMemo over HTTPS.", "Provides ranked, bounded context recall before a turn.", "Includes a local SQLite reliability layer." ], "readmeHash": "3a93ff56efd19a9f15a3b042a2ca7f6fafa0aab228f2276dc1c6bd0095245b69", "generatedAt": "2026-10-19T15:06:53.896Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "abundantbeing/hermes-browser-extension": { "summary": "Hermes Browser Extension is a Chromium side panel and Firefox sidebar for connecting browser context to the Hermes Agent runtime through a local gateway, Hermes Cloud, or a self-hosted remote gateway. It includes chat, attachments, page-context capture, and, in Chromium, optional Hermes Control for operating attached tabs.", "highlights": [ "Supports local, Hermes Cloud, and self-hosted remote connections.", "Bot Mode supports agent switching and group chats.", "Hermes Control is opt-in and Chromium-only." ], "readmeHash": "0d8d298b99fa8636b5fe25128969e749f53489246b9c18d1e8ccf772ef33198f", "generatedAt": "2026-10-19T15:06:21.534Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "tinyhumansai/tiny.place": { "summary": "tiny.place provides ainfrastructure layer for autonomous AI agents to establish identities, communicate, and conduct commerce. The system utilizes a registry for @handle identities, Signal Protocol for encrypted messaging, and Solana for on-chain payment settlement. It enables agents to discover one another through an open directory and transact using USDC and SOL via an x402-compliant facilitator.", "highlights": [ "Provides encrypted agent-to-agent messaging via Signal Protocol", "Supports identity registration with human-readable @handles", "Facilitates commerce through Solana-based USDC and SOL payments" ], "readmeHash": "55e8881df46a797380ca566ac8745ece58b618bcaed577c6f1c90c03e4ed1e07", "generatedAt": "2026-06-27T07:36:59.166Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:10:58.746Z" }, "observeco/observeco": { "summary": "ObserveCo is an open-source observability tool for monitoring and benchmarking AI agents. It provides a dashboard to track token usage, costs, and task success rates through passive traffic analysis and active probing.", "highlights": [ "Native Hermes plugin exports real-time telemetry from agent conversations.", "Two-mode monitoring includes zero-token passive analysis and scheduled active probing.", "Config-aware baselines detect performance drift relative to local file changes." ], "readmeHash": "0497c24bf72286f3453286c416c2f0725f79c1cb609cbc1a4ff9aa24f92ed229", "generatedAt": "2026-09-04T10:54:59.449Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-07T13:43:58.402Z" }, "DanielLi202/hermes-tag": { "summary": "Hermes Tag is a context-selection layer for Hermes on Feishu/Lark, Slack, and DingTalk. It retrieves specific images, notes, and relevant replies when @-mentioned to provide bounded evidence without processing entire chat histories.", "highlights": [ "Provides per-chat memory isolation", "Supports Feishu, Slack, and DingTalk", "Uses Tier-0 short-term and Tier-1 long-term memory" ], "readmeHash": "971c6507d9b388f485c3e7d641fc9d641fc83d6671759d74c368d4f981ea36e5", "generatedAt": "2026-07-14T21:39:14.141Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-14T21:42:41.309Z" }, "bergside/typeui": { "summary": "TypeUI is an AI-first platform that provides design skills, UI prompts, design systems, and layout variation guidance to help AI tools generate user interfaces for websites and applications. Its resources work with AI providers that support Markdown files or general prompting, and it connects to a range of coding tools through TypeUI MCP or plugins.", "highlights": [ "Provides design skills, UI prompts, and layout guidance.", "Supports AI providers that accept Markdown or general prompting.", "Connects to coding tools through TypeUI MCP or plugins." ], "readmeHash": "4ee455e4a0691faca1ce518804885daddcf612b109e1e287481340c93b7a30e1", "generatedAt": "2026-10-23T05:06:03.804Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Cyber-Yichen/hermes-xiaomi-mimo-tts": { "summary": "This plugin integrates Xiaomi MiMo V2.5 TTS into Hermes Agent to enable high-quality voice responses. It de-codes MiMo's Chat Completions API output and uses ffmpeg to transcode audio into OGG/Opus format. This process allows for native voice-message delivery within Feishu and Lark environments. The implementation avoids external dependencies like the OpenAI SDK and ensures secure credential handling via environment variables.", "highlights": [ "Converts MiMo TTS to OGG/Opus for Feishu voice messages", "Supports standard API and Token Plan credentials", "Automatic ffmpeg fallback to MP3 audio attachments" ], "readmeHash": "bad904a6a9fe1474b10412c1b62f31c21579e855010b634d358f4346fd4916d3", "generatedAt": "2026-07-02T07:52:58.382Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:11:13.764Z" }, "penfieldlabs/hermes-penfield": { "summary": "hermes-penfield is a persistent memory provider that enables Hermes Agent to maintain memory across different sessions and agents. It connects the agent to the Penfield memory API using a stdlib-only HTTP implementation with OAuth and API-key authentication. The provider utilizes hybrid search and a knowledge graph with 24 relationship types to manage memories and artifacts. It also automatically saves checkpoints of relevant memories when context compression occurs.", "highlights": [ "17 tools for memory management and knowledge graph traversal", "Hybrid search combining BM25, vector, and graph methods", "Automatic context checkpointing during message compression" ], "readmeHash": "82280b262eb4755019a501231cf7de255c77365687cf4159c72ea6b2b189b2ce", "generatedAt": "2026-07-02T07:54:23.082Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:11:16.602Z" }, "Cyber-Yichen/hermes-feishu-group-context": { "summary": "Hermes Feishu Group Context is a plugin that provides local archiving and on-demand context retrieval for Feishu/Lark group chats. It uses a local SQLite database to store messages, ensuring that model tokens are only consumed when the agent is explicitly triggered. The system automatically injects recent history during empty mentions or provides specific time-range summaries via a dedicated tool. This approach allows the agent to maintain group awareness without processing every message in real-time.", "highlights": [ "Local SQLite storage for group chat archiving and persistence", "Automatic context injection for empty mentions and relevant queries", "Customizable retention periods and context window sizes per group" ], "readmeHash": "5aa74e853b19d1fb51c56754b44b75d02ea762698a6935226634f3746290b293", "generatedAt": "2026-07-03T04:09:40.277Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:11:19.288Z" }, "aidevelopers2/remoteopenclaw-mcp": { "summary": "Remote OpenClaw MCP Server provides a command-line and agent-native interface for searching a directory of over 13,000 MCP servers and 4,000 agent skills. It functions as a thin TypeScript wrapper around the Remote OpenClaw public search API, allowing agents to retrieve tool metadata and installation commands without a browser. This integration enables agents like Hermes and Claude Code to autonomously discover and suggest relevant plugins for tasks such as database management or web scraping. The server requires no API key and integrates directly into standard MCP client configurations.", "highlights": [ "Searches over 13,870 MCP servers and 4,384 agent skills", "Provides direct installation commands for discovered tools and plugins", "Integrates with MCP-compatible clients like Hermes, Cursor, and Claude Code" ], "readmeHash": "ed28a86a07624c98b0c7993f3608b05817f2703912604d33b6dd0e520f7b49a9", "generatedAt": "2026-07-03T04:10:02.147Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:11:21.679Z" }, "chainbase-labs/Agentkey": { "summary": "AgentKey is a unified data access tool that provides AI agents with internet connectivity to social media, web pages, and on-chain data through a single subscription. It automates the configuration of external APIs and data scraping capabilities for over 40 supported agents, including Claude Code and Cursor CLI.", "highlights": [ "Provides access to Twitter, LinkedIn, Reddit, and GitHub data.", "Automates configuration for over 40 different AI agents.", "Integrates via a CLI-managed MCP loader for DeepSeek Harness." ], "readmeHash": "d7699012070123d704d75eebd65de37cbcfd20a2b9b8c2d9fb8d38f007ebca7a", "generatedAt": "2026-09-07T13:45:51.262Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-14T14:29:23.806Z" }, "teixeirazeus/fablize-for-hermes": { "summary": "fablize-for-hermes is a port of the fablize procedure system designed to improve task completion and verification within the Hermes Agent ecosystem. It implements a structured behavioral ruleset that enforces multi-step task decomposition, evidence-based checkpoints, and systematic investigation protocols for debugging. By utilizing a Python-based goal engine and specific verification packs, the project ensures agents ground their claims in tool results rather than ending tasks prematurely. This framework helps maintain procedural discipline during complex coding, UI rendering, and logic tasks.", "highlights": [ "Implements multi-story task decomposition with mandatory evidence-based verification gates.", "Provides a systematic investigation protocol for reproduction and causal debugging.", "Includes a Python goal engine for persistent task state management." ], "readmeHash": "261a679d6153a86257b0113a1292aa5f169cd270d8780bd6c68337cb553e215e", "generatedAt": "2026-07-03T19:11:09.359Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:11:29.185Z" }, "internet-court/internet-court-skill": { "summary": "Internet Court is an open skill designed for agent-to-agent commerce that integrates fragmented contacto protocols into a single lifecycle. It provides a framework for structuring deals, holding funds, and establishing adjudication for dispute resolution.", "highlights": [ "Provides a master router and connector skills", "Includes vendored copies of official protocol skills", "Adds a verification and dispute resolution layer" ], "readmeHash": "478f407a6e2c3b2e50fcb4f9852ff0886919472f3340282432deda999ebdc2e6", "generatedAt": "2026-08-20T07:13:43.928Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-24T08:33:28.431Z" }, "supernovae-st/nika": { "summary": "Nika turns repeatable AI tasks into readable `.nika` workflow files. It checks workflows before running them, lets users choose a model and run on their machine or through a provider API, and records runs in tamper-evident traces that can be verified.", "highlights": [ "Checks workflows before any model is called", "Includes a Rust binary and TypeScript SDK", "Open source under AGPL-3.0-or-later" ], "readmeHash": "c09a8257b1176cabc47aad6315f06e0a314d06bef5ff83104c9ad1297f04e948", "generatedAt": "2026-10-26T07:19:29.257Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "longsizhuo/openInvest": { "summary": "OpenInvest is a self-hosted investment decision engine designed to work with AI agents. It provides an investment committee with isolated analysis and debate, backtesting, and auditable decision records; the README says its purpose is transparency and discipline, not reliable directional prediction or excess returns.", "highlights": [ "Multi-agent committee with isolated analysis and rebuttal debate", "Auditable decision logs and long-horizon backtesting", "Agent skill and daily report deployment workflow" ], "readmeHash": "af37b21c0bfa434709249e56973de0c2535f664cfb6ec510b15bcadb8c2fefcd", "generatedAt": "2026-10-20T21:46:19.512Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "solomon2773/nora": { "summary": "Nora is a self-hosted platform for deploying and operating OpenClaw and Hermes agent fleets on Docker or Kubernetes. It provides lifecycle controls, logs, monitoring, secrets management, workspaces, and access through a web interface, REST API, CLI, and MCP server.", "highlights": [ "Supports Docker and Kubernetes deployments.", "Includes searchable logs, metrics, and alert rules.", "Provides REST API, CLI, and MCP server access." ], "readmeHash": "0977576f2de771ad1c52b0a45359fa449ce0efad1d059389c1ba8b40b0bb7ddb", "generatedAt": "2026-10-15T17:13:04.247Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:10.436Z" }, "Sibyl-Labs/Sibyl-Memory": { "summary": "Sibyl Memory is a local-first agent memory system with a SQLite-backed, five-tier schema and FTS5 search; the README says it uses no vector database or embedding model. Its five packages provide an SDK, command-line interface, Hermes integration, MCP server, and LangGraph BaseStore adapter.", "highlights": [ "Stores memory locally in a SQLite file", "Includes FTS5 search and a five-tier schema", "Offers SDK, CLI, Hermes, MCP, and LangGraph packages" ], "readmeHash": "dd00a953dd2a7cbb17d95dc08a1a1e98b28b299e5cd7f5fc22b244f0258dd97d", "generatedAt": "2026-10-15T04:55:46.336Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:12.347Z" }, "obra/superpowers": { "summary": "Superpowers is a software development methodology for coding agents, built from composable skills and initial instructions that guide agents to use them. Its workflow elicits and presents a specification, creates an implementation plan emphasizing TDD, YAGNI, and DRY, then uses subagents to carry out and review engineering tasks.", "highlights": [ "Skills trigger automatically in supported coding-agent harnesses.", "The workflow includes specification, planning, and subagent-driven development.", "The README lists installation instructions for multiple coding agents." ], "readmeHash": "79f29f50c735862d2e304ea547b4c9aa67d9db6909ff6fb7b5ea922d7bcbc3e7", "generatedAt": "2026-10-20T21:46:19.583Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Capslockb/hermes-live-discord-agent-plugin": { "summary": "Hermes Live is a self-hosted Hermes Agent plugin that connects Discord voice channels to the Google Gemini Multimodal Live API. It enables real-time voice conversation and supports optional tool integrations and notifications.", "highlights": [ "Full-duplex Discord voice input and output", "Supports Gemini Multimodal Live sessions", "Provides JSONL voice-event records" ], "readmeHash": "dd5f8aa9cf8354a01c0ed6116d8059c42e14d80099303961f1614a3e506ba09e", "generatedAt": "2026-08-04T00:59:26.248Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-10T08:29:47.849Z" }, "indranilbanerjee/contentforge": { "summary": "ContentForge is an open-source content production plugin that runs a 10-phase pipeline to produce publication-ready .docx files with fact-checking, internal linking, and C2PA provenance. It includes a humanizer, a run auditor that re-derives quality gates, and a lifecycle loop for measuring and auditing published content. It installs on Claude Code, Cowork, Codex, Cursor, Copilot CLI, and other agent platforms.", "highlights": [ "10-phase pipeline with script-checked quality gates", "Outputs .docx with C2PA provenance sidecar", "Requires typed yes before publishing or translating" ], "readmeHash": "2c4a34f5e14f2a557ae183c9467e59d97342e40e45cc675d0a2c4827bab82138", "generatedAt": "2026-10-12T00:27:07.129Z", "model": "deepseek/deepseek-v4.1-flash", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:04:47.480Z" }, "piprail/piprail": { "summary": "PipRail is an implementation of the x402 \"402 Payment Required\" standard designed for agent-to-merchant payments. It provides an SDK and MCP server to enable direct wallet-to-wallet transactions across EVM, Solana, TON, Tron, NEAR, Sui, Aptos, Algorand, Stellar, and the XRP Ledger.", "highlights": [ "Provides a TypeScript SDK for x402 payments", "Includes an MCP server for budget-bound agent wallets", "Operates without a backend, database, or API keys" ], "readmeHash": "5f77abc32850a2c21b2d74e3ca83daff12270669f2ae46f3eb38229fe8f585af", "generatedAt": "2026-09-14T14:32:04.022Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:36:47.379Z" }, "ccf/agentcairn": { "summary": "agentcairn provides a cross-project memory system for coding agents that stores context as Markdown files in a local vault. It uses a disposable DuckDB index for retrieval and supports integration with tools like Claude Code and Codex via MCP plugins.", "highlights": [ "Stores durable memory as inspectable Markdown files with frontmatter metadata.", "Uses a local DuckDB cache for hybrid search and retrieval.", "Integrates with coding agents via MCP servers and host-specific plugins." ], "readmeHash": "c90874b2ba07bbd9f6f2044cf30520b213f39ea29411aa20d5060eaee97410d9", "generatedAt": "2026-09-14T14:32:05.508Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:36:49.698Z" }, "sourcevault-ai/sourcevault-code-tools": { "summary": "sourcevault-code-tools is a Hermes Agent plugin for searching code, reading exact files, and answering repository questions with file-and-line citations. It connects to a separate SourceVault backend and is described as running on the user’s machine over localhost without source code leaving it.", "highlights": [ "Provides slash commands for search, file reads, context, and Q&A.", "Supports Hermes CLI, Telegram, and Hermes Desktop.", "Includes repo sync and indexed git-history search commands." ], "readmeHash": "bcd126e7054f4e5c58fe8e6ea0419a479a64ac276ab743a11ee865a8585c7505", "generatedAt": "2026-10-12T17:07:34.720Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:13.952Z" }, "Shine8592/china-briefing": { "summary": "China Briefing is a multi-scale Chinese news briefing generator designed for Hermes Agent. It generates structured reports ranging from global to street-level geographic scales using Tavily API for web searches.", "highlights": [ "Supports multiple geographic scales from global to street level", "Generates structured reports with core conclusions and action suggestions", "Integrates user interests via semantic memory for search expansion" ], "readmeHash": "4e39a5c8b1e6a7274d1c07d6068c4942ec17a4fae9119b5f18e579b11f52b599", "generatedAt": "2026-07-16T21:10:03.196Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-16T21:19:30.137Z" }, "IVRZ-da/agentiker-code-intel": { "summary": "agentiker-code-intel is a Hermes plugin providing AST-aware code intelligence using tree-sitter, ast-grep, and LSP. It offers 70 tools for tasks such as symbol extraction, structural search, and impact analysis across supported languages.", "highlights": [ "Supports C, C++, Go, Java, JavaScript, Python, Rust, TSX, and TypeScript", "Uses a SQLite-persistent import graph for knowledge queries", "Provides LSP-powered definition, references, and diagnostics" ], "readmeHash": "bf31ca30ac17ab9fe26296825eeb1ddfff0dffe0810ffee5b5f6f5fe34e21eb4", "generatedAt": "2026-07-16T21:22:05.996Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-20T09:00:38.150Z" }, "corsur/swarm-tips": { "summary": "Swarm Tips is an AI agent platform featuring Solana and EVM smart contracts for decentralized coordination and task marketplaces. It includes the Coordination Game for anonymous social deduction and Shillbot for AI-driven content creation tasks, supported by a Model Context Protocol (MCP) server for agent interaction.", "highlights": [ "Supports Solana (Anchor) and EVM (Foundry) smart contract deployments.", "Provides MCP server endpoints for AI agent discovery and messaging.", "Implements on-chain escrow for task payments and game stakes." ], "readmeHash": "33609bdfcb8a4dce7e5e1b41a7bc267c391148934fed76bd97830bc988d2f156", "generatedAt": "2026-09-21T05:10:42.250Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:36:50.792Z" }, "TauricResearch/TradingAgents": { "summary": "TradingAgents is a multi-agent framework that models the dynamics of real-world trading firms. It assigns tasks to specialized roles, including analyst, trader, and risk-management roles, and supports dynamic discussions about trading strategies. The README says it is designed for research purposes and is not financial, investment, or trading advice.", "highlights": [ "Uses specialized roles for trading tasks", "Agents engage in dynamic strategy discussions", "Designed for research purposes" ], "readmeHash": "fa3c3b517208e804b880c447f72a9f4730d856bc2f414d4618cf41e8c44d0cd7", "generatedAt": "2026-10-19T15:06:57.805Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "IVRZ-da/agentiker-plan-follow": { "summary": "plan_follow is a Hermes plugin for structured plan creation and task enforcement. It provides tools for managing task dependencies, parallel execution groups, review gates, and git integration.", "highlights": [ "Supports 12 templates for plan creation", "Includes 6 review profiles for task validation", "Provides git integration for branches, commits, and PRs" ], "readmeHash": "2741081a17f761d03df65827843940089ce864044c9fd336e94e8b37270c4b2e", "generatedAt": "2026-07-16T21:22:07.172Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-20T09:00:50.475Z" }, "IVRZ-da/agentiker-scout": { "summary": "agentiker-scout is a Hermes plugin providing tools for code and architecture analysis, vulnerability scanning, and autonomous web research. It utilizes a shared pattern pipeline to allow findings from analysis scans to inform automated bug-hunt scans.", "highlights": [ "Includes 56 tools across three functional domains", "Supports automated pattern-based vulnerability scanning", "Provides autonomous web research with lifecycle management" ], "readmeHash": "7ab80ab5d5cf4a36d8c81d6dc1c2e301c9c8c4dca3d02e1d79f5c31d1f566734", "generatedAt": "2026-07-16T21:22:11.481Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-20T09:00:49.310Z" }, "visorcraft/MongrelDB-Hermes": { "summary": "MongrelDB-Hermes is a hybrid long-term memory plugin for the Hermes Agent that uses multiple index types for retrieval. It supports dense ANN, sparse retrieval, exact text, and metadata filtering with built-in MinHash deduplication.", "highlights": [ "Supports AES-256-GCM encryption at rest by default", "Offers native Rust FFI or HTTP daemon execution modes", "Includes HNSW ANN, sparse, and FM-index retrieval" ], "readmeHash": "d8ab931e102837007ad15c7f75f4607cc7abf52c1aa6ba3ebe8f2531e2310acc", "generatedAt": "2026-08-12T02:51:05.740Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-17T08:29:06.334Z" }, "Fruxano/fruvisi": { "summary": "FruVisi is a dashboard plugin for Hermes Agent that provides an interactive 2D organizational chart for managing agent structures. It enables functional presets for team lineups, per-agent OpenRouter fallbacks, and kanban task integration via the native Hermes API.", "highlights": [ "Interactive 2D org chart using React Flow", "One-click functional presets for team lineups", "Per-agent OpenRouter fallback configuration" ], "readmeHash": "80bd0cd31fa9572965812becd720234bc35457744266250455a23d450e109d52", "generatedAt": "2026-07-20T15:31:39.081Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-27T09:03:26.394Z" }, "stepanov1975/hermes-local-knowledge": { "summary": "Hermes Local Knowledge is a reusable Hermes Agent plugin that routes local questions to likely artifacts, such as skills, documents, scripts, runbooks, cron jobs, MCP servers, and generated tool-routing notes. It provides tools to search, fetch, and traverse artifacts, and its search results are routing hints; the agent should read the selected source before acting.", "highlights": [ "Registers five tools for search, retrieval, traversal, feedback, and usage reports.", "Indexes configured local sources and selected Hermes runtime sources.", "Registers pre_llm_call, on_session_end, and on_session_finalize hooks." ], "readmeHash": "a149d21a80f2ca10bfc52820f4c26fa3a733560fe7c580a5aeb4eb2eeda62eca", "generatedAt": "2026-10-12T17:07:39.163Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:17.089Z" }, "chenwei791129/hermes-usage-hook": { "summary": "Hermes Usage Hook appends a line showing rate-limit usage to each reply when it recognizes a supported provider from the reply’s model name. It supports Codex and MiniMax, skips the footer if fetching usage fails, and requires the relevant provider credentials.", "highlights": [ "Shows usage for available 5-hour and weekly windows.", "Supports Codex and MiniMax model names.", "Failed usage fetches do not break replies." ], "readmeHash": "2f5b7e87066618c6e3cc92b52afa8c4a70a6c45a634576cbb7930330123ba614", "generatedAt": "2026-10-23T15:08:43.879Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "ClaudioDrews/memory-os": { "summary": "Memory OS is a local memory infrastructure for Hermes Agent that utilizes a seven-layer system to store and retrieve information across sessions. It integrates SQLite, Qdrant, and Redis to provide structured fact storage, semantic search, and an auto-curated wiki pipeline while remaining compatible with various LLM providers.", "highlights": [ "Seven-layer memory system including vector search and structured facts", "Local infrastructure using Docker, Qdrant, Redis, and SQLite", "Ground Truth hierarchy to ensure agent prioritizes injected context" ], "readmeHash": "bce8f8e1a41797b9caff8d283ecd769559a00c49e6a17e8e969c70b57d35144c", "generatedAt": "2026-07-28T23:43:01.645Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-29T14:19:19.198Z" }, "uzairansaruzi/hermex": { "summary": "Hermex is a native SwiftUI iPhone app that connects to a self-hosted hermes-webui server to control an AI agent running on hardware you manage. It supports streaming chats, session browsing, model and profile selection, scheduled tasks, skills, and server-file browsing.", "highlights": [ "Stream chats and attach files or images", "Browse, search, and resume server sessions", "View scheduled tasks, skills, memory, and insights" ], "readmeHash": "f2f077e75d9a15729cfeeff083506260fb82695e9c163cd877791f90fd961cb7", "generatedAt": "2026-10-22T05:07:56.467Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "sharbelxyz/hermes-agent-mission-control": { "summary": "Hermy HQ is a self-hostable dashboard template designed to manage a local Hermes AI agent through a shared Postgres database acting as a message bus. Users can dispatch tasks, review agent plans, and manage memory via a web interface while the agent remains on a local machine.", "highlights": [ "Uses a Postgres message bus to bridge local agents and web dashboards", "Features an approval inbox for requests with side effects", "Built with Next.js 16, React 19, Prisma, and Tailwind CSS v4" ], "readmeHash": "c43cd0a20124bb3eafe5757ff93060892835ddabcff27e6b21fdbd39a554a74a", "generatedAt": "2026-07-28T23:43:23.635Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-29T14:19:25.286Z" }, "bookunt3d/hermes-agent-fa": { "summary": "This repository contains a complete Persian translation of the official Hermes Agent documentation. The project uses GitHub Pages and includes full support for right-to-left (RTL) text direction.", "highlights": [ "Complete Persian translation of Hermes Agent documentation", "GitHub Pages site with RTL support" ], "readmeHash": "b5767ce8e8b192ea7b380a0f6d584f083fa32d342ca7aa0a39fcaff96cde29e7", "generatedAt": "2026-07-28T23:43:27.548Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-29T14:19:28.354Z" }, "amirghm/hermes-agent-mobile": { "summary": "This project enables the Hermes AI agent framework to run on Android via Termux and iOS via iSH. It allows users to interact with multiple AI personas through a CLI or via Telegram using an API key.", "highlights": [ "Supports Android (Termux) and iOS (iSH)", "Integrates with Telegram for voice and file messaging", "Supports scheduled tasks via cron jobs on Termux" ], "readmeHash": "f29430d855522c0ad51d9a3165feaa595e6552d769ec3aa716665d74a45edade", "generatedAt": "2026-07-28T23:43:28.110Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-29T14:19:28.347Z" }, "InfiniteWhispers/HermesAgent-MultiModel": { "summary": "HermesAgent-MultiModel is a local-first AI agent framework that uses Mixture-of-Agents (MoA) orchestration to route tasks to specialized models. It integrates with Hermes and Ollama to run agentic workflows on consumer GPUs without cloud dependencies.", "highlights": [ "Implements Mixture-of-Agents via parallel model querying and synthesis", "Uses Hermes for task routing and tool-calling", "Optimized for RTX 4080/5080 hardware" ], "readmeHash": "a35df71ea2a928f953402bdfe36ab312dd01ff2c2db98fbde3f20f41fd2e77d2", "generatedAt": "2026-07-28T23:43:32.922Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-29T14:19:31.299Z" }, "0xarkstar/awesome-hermes-agent": { "summary": "This project is a curated list of resources, tools, and projects for Hermes Agent, a self-improving AI agent by Nous Research. It includes links to official documentation, deployment options, skills, and various community integrations.", "highlights": [ "Curated list of Hermes Agent resources", "Includes deployment and skill integration links", "Links to official Nous Research documentation" ], "readmeHash": "8461965c5af0e01db0870c3e859977f377d975399dd41b267585c6dd347e8e73", "generatedAt": "2026-07-28T23:43:35.696Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-29T14:19:31.619Z" }, "zcweah1981/awesome-hermes-agent-zh": { "summary": "This project provides a Chinese-language documentation hub and implementation guide for Hermes Agent. It offers structured paths for local deployment, domestic model integration in China, and migration from OpenClaw.", "highlights": [ "Guides for integrating DeepSeek, Alibaba, Tencent, and other domestic models.", "Documentation for multi-agent orchestration and MCP tool integration.", "Troubleshooting references for installation, CLI, and gateway issues." ], "readmeHash": "435cd85f4f97a5739a535c92d5182c3f4730871ece4aa22b3692f500597d479a", "generatedAt": "2026-09-30T22:30:47.185Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:04:55.155Z" }, "Javis603/token-monitor": { "summary": "Token Monitor is a desktop app that displays live token usage and AI tool limits across more than 47 AI coding tools, including Claude Code, Codex, and Hermes Agent. It offers historical usage trends and breakdowns by tool, device, model, session, or project, with multi-device sync and views in the main window, menu bar or system tray, and other desktop locations.", "highlights": [ "Supports token usage, tool limits, and session details.", "Provides historical trends and usage breakdowns.", "Offers real-time multi-device sync." ], "readmeHash": "54bdab496abdba808d0c6e83f4125f3b35e3cf9097ba46aa3a708d612f81d0b1", "generatedAt": "2026-10-25T16:35:46.924Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "freestylefly/wesight": { "summary": "WeSight is an open-source desktop control console for local AI agents, providing a visual workspace to manage coding agents, tool execution, and file changes. It supports the installation and configuration of multiple agent engines and model providers through a unified graphical interface.", "highlights": [ "Supports Claude Code, Codex, OpenClaw, and Hermes Agent engines.", "Includes a runtime dashboard for tracking token usage and latency.", "Provides a live workspace for monitoring code writes and diffs." ], "readmeHash": "c7e073f1958ea9ccd08060830fac79eb8eaba01e082cc0a8646fab5b63e7bb83", "generatedAt": "2026-07-28T23:43:44.139Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-29T14:19:37.097Z" }, "spacering-net/codeg": { "summary": "Codeg brings multiple AI coding agents into one workspace, where they appear in a shared structured conversation interface. It can import, search, and resume agent histories, hand sessions to another agent, and support cross-agent delegation and queued to-dos in separate worktrees. The README lists desktop, self-hosted server, Docker, mobile, and messaging-app access.", "highlights": [ "Includes 19 agents and supports ACP agents.", "Import, search, resume, and hand off agent sessions.", "Delegate across agents or queue to-dos in worktrees." ], "readmeHash": "472c9ad6473d6476b802829e704472af8ea375577e902557e2e46e5cb1406cba", "generatedAt": "2026-10-19T04:55:53.192Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:19.734Z" }, "Eynzof/Hermes-CN-Desktop": { "summary": "Hermes Agent CN Desktop is a desktop client for the Hermes Agent Chinese community, built with Tauri v2, Rust, React, and TypeScript. It provides a native interface for Windows and macOS to manage the Hermes-CN-Core agent runtime, local processes, and model service configurations.", "highlights": [ "Supports Windows and macOS with native window and process management.", "Features a built-in Hermes-CN-Core kernel with health check and update capabilities.", "Includes UI for chat, memory management, and LaTeX/Markdown rendering." ], "readmeHash": "2ba16ed98e2c3fd84b801f2419ebe1ab2feed310305b3b3468fca97b580065e1", "generatedAt": "2026-09-07T13:45:27.052Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-14T14:29:38.110Z" }, "agent37-platform/minions": { "summary": "Minions is a centralized management interface for Hermes Agent that replaces fragmented terminal sessions with a unified dashboard. It utilizes a Kanban-style board to track persistent agent sessions, allowing users to monitor reasoning, tool calls, and task progress in real time. The system enforces a human-in-the-loop workflow where agents propose completions that require manual verification before being marked as done. Users can also manage scheduled recurring jobs and browse agent-created files through a local-first SQLite architecture.", "highlights": [ "Kanban board for tracking autonomous task status and review queues", "Live streaming of agent reasoning and tool execution", "Human-in-the-loop verification for all agent-proposed task completions" ], "readmeHash": "1d0a00a8a089656c915ab2398a475aae14c83549aa214e58d977cb056105b32f", "generatedAt": "2026-05-28T12:12:37.315Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:08:23.245Z" }, "hermes-labs-ai/lintlang": { "summary": "LintLang is a local, deterministic static linter for AI agent instructions and tool interfaces. It scans supported files for issues such as ambiguous tools, missing bounds, schema gaps, and prompt defects, and can emit JSON, SARIF, and GitLab Code Quality reports.", "highlights": [ "Scans agent-facing content in project directories.", "Findings include suggested fixes and file locations.", "Supports CI gates and reviewed baselines." ], "readmeHash": "60e62b9eebe6a769cba8eeaa707a318b5efc75bb333b06601ad83857bc54ce54", "generatedAt": "2026-10-21T01:14:35.157Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "runta-dev/clawshell": { "summary": "ClawShell is a security-privileged process that acts as a safety harness for the Hermes Agent ecosystem by protecting sensitive credentials and PII. It functions as a proxy between the agent and upstream LLM providers, swapping virtual keys for real ones stored in a restricted directory to ensure the agent never has direct access to credentials. The system performs Data Loss Prevention (DLP) scanning to redact or block sensitive information in request and response bodies. Additionally, it provides secure email isolation through IMAP sender filtering and supports OAuth-based authentication for providers like OpenAI.", "highlights": [ "Maps virtual keys to real credentials stored in protected Unix directories", "Scans and redacts PII in request and response bodies via regex", "Enforces sender-based email filtering for secure mailbox access" ], "readmeHash": "a924345a16e04c033f9fcc3fbe986b4bd4db4c6837c14b4f1600c46857d2e4d1", "generatedAt": "2026-04-17T18:20:48.096Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:06:08.522Z" }, "EKKOLearnAI/ekko-studio": { "summary": "Ekko Studio is a local-first AI workspace available as a desktop app and self-hosted web console. It supports Hermes Agent, Ekko Agent, and coding-agent integrations for chat, workflows, files, voice, and device tools, with session management and platform-channel configuration.", "highlights": [ "Supports multi-agent chat, group chat, and visual workflows.", "Includes session management and usage details.", "Ships as a desktop app, npm CLI, and Docker image." ], "readmeHash": "c5938f475da99a617eec6b54187e220848aa354dbbb43e5c205e005ec3c8e6fa", "generatedAt": "2026-10-26T05:02:23.745Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "generalbusiness-ai/keep": { "summary": "Keep is a reflective memory system for AI agents that facilitates long-term learning and contextual awareness. It processes diverse inputs—including URLs, codebases, and multimedia—by summarizing, embedding, and organizing them into a searchable graph. The system uses 'edge tags' to create navigable relationships between notes, such as linking speakers to transcripts or commits to files. It integrates directly with Hermes Agent and other MCP-compatible tools to help agents track intentions and reflect on past actions.", "highlights": [ "Semantic search with BM25 full-text and ranked graph traversal", "Automated indexing for codebases, URLs, and multimedia with OCR support", "Edge tags create navigable, bidirectional relationships between stored notes" ], "readmeHash": "d2d59978d9d90c5c1fe5105e62663d5fe0f09f925dbca325e659becaeecb8ea0", "generatedAt": "2026-07-06T09:52:25.185Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-07-13T09:06:52.271Z" }, "mnemosyne-oss/mnemosyne": { "summary": "Mnemosyne is a Python memory layer for AI agents, using SQLite and requiring no external services. It offers a Python API, a command-line interface, and MCP support, with integrations documented for several agent platforms.", "highlights": [ "Stores and recalls memories in SQLite.", "Supports MCP clients and direct Python use.", "Includes CLI commands for memory operations and sync." ], "readmeHash": "f6c0f72c1289cf5f1ce974dc144b800ab868106f607db8bbd16f7187eca3bfd8", "generatedAt": "2026-10-25T16:35:38.940Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "chillerno1/hermes-yt-plugin": { "summary": "This plugin provides a compact YouTube player for the Hermes Desktop client that functions in floating or docked modes. It uses an isolated Electron webview to play watch pages, playlists, Shorts, Live, and embed links without requiring an API key.", "highlights": [ "Supports floating or docked pane modes", "Saves favourites and up to 50 recent videos", "Uses an isolated webview session" ], "readmeHash": "c3e2217ec6379b211dd4a7b91f442a122e60be4f43cfa826bafd174bbb2eb70e", "generatedAt": "2026-07-30T11:49:11.101Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-03T09:02:49.909Z" }, "topoteretes/cognee": { "summary": "Cognee turns documents, code, and conversations into a self-hosted knowledge graph that agents can search and reuse across sessions. Its local quickstart builds and queries memory using local extraction and embedding models without an LLM; generated answers require additional LLM configuration.", "highlights": [ "Structures text as entities, relationships, and searchable chunks.", "Supports custom data models and ontologies.", "Offers plugins and MCP for connecting agents." ], "readmeHash": "a19c715a2c161960f222eb1d52b1f28048bf6d998c8e7b7013bc3cf024c15a89", "generatedAt": "2026-10-19T15:07:01.331Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "topoteretes/cognee-integrations": { "summary": "Cognee Integrations is a monorepo of Cognee-owned integration packages that give agent frameworks a persistent memory layer backed by cognee, described as a permanent knowledge graph plus a fast session cache. It lists installable packages for frameworks including Strands, CrewAI, LangGraph, Google ADK, Claude Agent SDK, Hermes Agent, OpenClaw, n8n, and Dify, plus chat bot and editor integrations that talk to a running cognee server over its HTTP API. A Claude Code plugin provides persistent memory across sessions by auto-capturing prompts, tool traces, and responses and auto-recalling context.", "highlights": [ "Monorepo of Cognee-owned integration packages", "Two tiers: permanent knowledge graph and session cache", "Claude Code plugin auto-captures and auto-recalls context" ], "readmeHash": "47104e80472b66d0458f5f7dbaa417bbdf7fc173d68cbf3ee6ab39c5812d1e85", "generatedAt": "2026-10-12T00:27:22.195Z", "model": "deepseek/deepseek-v4.1-flash", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:05:03.683Z" }, "Hermes-brasil/hermes-brasil": { "summary": "Hermes Brasil is a Portuguese-language community repository for Hermes Agent, with skills, guides, and resources for Brazilian users. It includes instructions for installing skills, recommended guides, and contribution steps; its README says integrations and community directories are planned, not part of the current repository.", "highlights": [ "Portuguese-language skills in SKILL.md format", "Guides cover installation, production, cron, MCP, and skill authoring", "Contributions are submitted through pull requests" ], "readmeHash": "1d6298c5dc4de3083d9bf1d4354ea6be9918f985c232cd411ae123032037c225", "generatedAt": "2026-10-19T13:40:38.945Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:27.545Z" }, "freehul/progressive-skill": { "summary": "Progressive Skill reduces the size of an agent’s skills index by demoting selected categories to compact count lines and applying a configurable character budget. It uses toolset mappings and recency-weighted skill usage to decide what remains fully visible; the README says the repository is no longer actively updated and is retained as a design reference.", "highlights": [ "Includes a Hermes plugin and an agent-agnostic Python decision core.", "Demoted skills remain discoverable through skills_list.", "Usage scores decay based on recency." ], "readmeHash": "a05a5836db12209b535c3e1604f7fd49ed66fdc5d5242dda3e40ab64960042c0", "generatedAt": "2026-10-15T04:55:51.735Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:25.892Z" }, "handnewb/hermes-voice": { "summary": "Hermes-voice is a voice-enabled personal assistant designed for Hermes Agent. It operates entirely locally using speech recognition, voice activity detection, wake word detection, and speech synthesis.", "highlights": [ "Supports 35 languages via Piper and 9 via Kokoro.", "Features adaptive turn-taking and barge-in capabilities.", "Runs offline with no API keys or accounts required." ], "readmeHash": "d421ae68bb89e0adc4f61cedbe50ad08973feabafbff76c864e913cfd0d72b97", "generatedAt": "2026-08-07T19:26:40.736Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-10T08:30:07.524Z" }, "MilkyWay008/Hermes-OTG": { "summary": "Hermes OTG is a portable, zero-install version of the Hermes Agent designed for Windows 10 and 11. It operates from any drive or folder using relative paths to avoid host system pollution and can run side-by-side with an installed Hermes instance using a separate port and data directory.", "highlights": [ "Runs from USB or any drive without host installation or registry writes.", "Includes Desktop, TUI, and Gateway modes with isolated data environments.", "Bundles Python 3.12, PortableGit, and Sysinternals rescue tools for offline use." ], "readmeHash": "62c86ac229d3ff6c6ee31e633ba39dee898a5dbc568f8a60f1fb8c985249640b", "generatedAt": "2026-08-17T08:32:03.416Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-24T08:33:50.514Z" }, "jiawood2006/hermes-skills": { "summary": "Hermes Skills is a collection of nine open-source tools designed for Chinese-language agent tasks such as video transcription, document OCR, and text refinement. The project provides scripts for processing Douyin links, generating e-commerce product scenes, and managing long-form writing memory.", "highlights": [ "Transcribes Douyin videos to structured Markdown and SRT subtitles.", "Removes common Chinese AI writing patterns using a rule-based engine.", "Extracts structured data from PDFs and images using macOS Vision." ], "readmeHash": "9c69377fac4528c3d09d78022c5ddfd0c24df3719f22da1639df6b81ea97c9f4", "generatedAt": "2026-09-19T16:00:19.168Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:37:07.565Z" }, "handnewb/lock-on-absence": { "summary": "Lock on Absence is a Python-based tool that automatically locks a workstation when the user is no longer present. It uses webcam frames and various recognition backends to determine presence and manage workstation state.", "highlights": [ "Supports LBPH, SFace, and LLM recognition backends.", "Includes a watchdog to lock the screen if the agent fails.", "Provides a FAR/FRR measurement harness and SIEM audit trail." ], "readmeHash": "fb53f99219e221c24db9162166848c24507d215b3938e72052376bd4b1b6e6d4", "generatedAt": "2026-08-11T16:52:50.148Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-17T08:29:50.150Z" }, "skillseal/skillseal": { "summary": "SkillSeal is a trust system for AI agent skills that uses verifiable evidence and cryptographic signatures to certify them. It employs an evaluator to test skills in isolated Docker containers and a certifying authority to sign the resulting evidence.", "highlights": [ "Tests structure, functionality, and security in isolated Docker containers", "Produces cryptographically verifiable certificates using Ed25519 signatures", "Provides a public API for certificate verification and skill submission" ], "readmeHash": "8328ba994dd9e2ee0582ecbfd816a96c1f1650679d07a4161f20c0d115f3e622", "generatedAt": "2026-08-11T16:52:52.547Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-17T08:29:58.050Z" }, "handnewb/hermes-cybersec-lab": { "summary": "Hermes Cybersecurity Lab is a turnkey ecosystem designed for the Hermes Agent. It provides a collection of 2,077 skills, 131+ tools, and 28 cybersecurity frameworks across eight repositories.", "highlights": [ "Includes 2,077 total skills across multiple repositories", "Installs 131+ tools for network, forensics, and exploitation", "Integrates 28 cybersecurity frameworks and standards" ], "readmeHash": "d02ed3f1be6f3d8053daf5a36a0ef4c59d767be8c2e2370c3fd597a5ab06b819", "generatedAt": "2026-08-11T19:05:40.219Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-17T08:29:58.157Z" }, "EntangledQuantum/Life_OS": { "summary": "Life OS is a personal app where an AI agent creates and maintains habits, schedules tasks, tracks goals, and adjusts plans based on what happened. Users tap to complete items, while data stays in a local SQLite file with no account or cloud sync.", "highlights": [ "Habits recur and are scored daily; tasks cover other activities.", "Includes an Android client and connects to agents over MCP.", "Runs locally and stores data in SQLite." ], "readmeHash": "79615bdbe2c4a0b762554ad289d0dcfe1dac0f2e7030aedf0ca3e86bb0abbd24", "generatedAt": "2026-10-20T11:51:27.622Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "CorsenAI/hermes-windows-runtime-skills": { "summary": "This community-maintained skill pack for Hermes Agent provides deterministic path routing and Python runtime selection across Windows, Git Bash/MSYS, and WSL. It enables agents to identify execution domains and verify existing Python interpreters without installing new software or modifying project files.", "highlights": [ "Routes file operations using environment-specific path syntax.", "Selects and verifies existing project-compatible Python interpreters.", "Supports native Windows, Git Bash/MSYS, and WSL environments." ], "readmeHash": "2d6c963e0bae3516847b0e1d073fecfb1d359dff05449efe3b629276d6a85818", "generatedAt": "2026-08-17T19:07:07.877Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-24T08:33:53.524Z" }, "CorsenAI/hermes-connector": { "summary": "Hermes Connector is an open-source Chrome extension that links a selected Hermes profile and session to Chrome tabs you choose. From Chrome’s side panel, you can attach tabs and let Hermes read or control them; version 0.2.4 can also discover Hermes Desktop sessions.", "highlights": [ "Only attached tabs are available to the selected Hermes session.", "Supports page interaction, tab controls, and requested screenshots.", "The extension and companion must use matching versions." ], "readmeHash": "e76f6c49fe471447984565216fe5629e21e953d48c3c489056c077a4cb305af3", "generatedAt": "2026-10-12T17:07:45.152Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:29.438Z" }, "webmilmind1/hermes-bounty-board": { "summary": "hermes-bounty-board is a Hermes Agent plugin for answering customer support bounties or running a bounty board. It uses x402 over HTTP; a business human approves the winning answer, and the submitting wallet receives 85% of the reward automatically. This Python plugin pays on Base.", "highlights": [ "Lists open bounties and provides public wallet earnings records.", "Lets agents buy ticket context and submit answers for review.", "Board creation returns a URL, API key, and USDC deposit addresses." ], "readmeHash": "1f4f06cbaac146a985f08368554ecfd9580d74950cafb5138d67592a848fe984", "generatedAt": "2026-10-12T17:07:45.931Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:29.816Z" }, "scavio-ai/hermes-agent": { "summary": "Scavio for Hermes Agent provides structured web data from 31 platforms and an Extract endpoint via a hosted MCP server. It includes five ready-to-use skills and is compatible with clients following the Agent Skills standard.", "highlights": [ "Exposes 191 tools via MCP server", "Includes five production-grade skills", "Supports Claude, Cursor, Goose, and OpenHands" ], "readmeHash": "69ab6ee6685f90660797ad91defbfb2fe2f903826dda5705f41848f17a0218d3", "generatedAt": "2026-08-19T08:50:48.665Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-24T08:33:56.344Z" }, "Heybinshao/hermes-appearance-hub": { "summary": "Hermes Appearance Hub is a Hermes desktop plugin that gathers appearance and conversation-behavior settings in a status-bar panel. It includes theme and language controls, paper texture simulation, LXGW WenKai font support, interface scaling, and other settings that take effect immediately and sync with the official settings page.", "highlights": [ "Provides a floating appearance panel opened from the status bar.", "Includes 14 themes, language controls, paper texture, and font toggles.", "Removes injected changes and restores official settings when disabled or removed." ], "readmeHash": "ec78439a54ca42faebae82c86637d86914c81768bbfe5b8996fc2dc200ef8711", "generatedAt": "2026-10-23T05:06:11.860Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "pomponchik/incontext": { "summary": "incontext is a Hermes Agent plugin that dynamically budgets output space against the active context boundary. It inserts an output cap when remaining space meets a configured reserve or caller-supplied limit, otherwise leaving the request unchanged to avoid truncated responses.", "highlights": [ "Budgets output space against Hermes' active context boundary.", "Uses a two-stage preflight and middleware policy for token management.", "Supports automatic context compression and provider-aware token counting." ], "readmeHash": "504808f14f7aad92d570f8302a70740a97d38561838e3e4186660547a07ef9c3", "generatedAt": "2026-08-22T02:32:38.050Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-24T08:33:57.097Z" }, "conorbronsdon/agent-skill-builder": { "summary": "agent-skill-builder is a tool for generating and validating Claude Code and portable Agent Skills using plain-language descriptions. It includes a runtime-aware validator that detects structural drift, unsafe commands, and credential-shaped strings while enforcing specific character limits for each profile.", "highlights": [ "Generates Claude Code extensions and portable Agent Skills core files.", "Validates skills against runtime-specific profiles with CI-friendly exit codes.", "Monitors official frontmatter specifications weekly to detect field drift." ], "readmeHash": "c465d018d1fb1c3c90a0d959503890b9a0451ff4144a661c38dc571910d657b8", "generatedAt": "2026-09-24T12:02:40.762Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-28T16:09:15.174Z" }, "conorbronsdon/agent-context-os": { "summary": "Context OS keeps durable project context and workflows in a Git-backed repository of Markdown files. Its adapters let supported coding agents use shared repository state, while reviewed setup, checkpoint, and close requests produce proposals and receipts.", "highlights": [ "Supports adapters for Claude Code, Codex, and other listed agents.", "Includes workflows for migration, handoffs, and memory proposals.", "Integrations are opt-in and are not enabled automatically." ], "readmeHash": "e12cecf75080f7de41a75c431d16ef8aed02fc8dabcbdfea780954c92d972469", "generatedAt": "2026-10-15T11:48:32.082Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:33.428Z" }, "conorbronsdon/demo-gif-skill": { "summary": "This project is a skill for Claude Code and agentskills.io-compatible agents that automates the creation of reproducible demo GIFs for repositories. It generates recording scripts using vhs for terminal interfaces or Playwright for web applications, then renders, optimizes, and embeds the resulting GIF into the README.", "highlights": [ "Generates reproducible vhs .tape or Playwright recording scripts.", "Optimizes GIF files to target sizes under 8 MB.", "Supports terminal CLIs, TUIs, web apps, and libraries." ], "readmeHash": "9d9b99e17834e94f4a22dc2363b83de03647802e841f612a8e21cf09b1257e46", "generatedAt": "2026-09-24T12:02:41.811Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-28T16:09:17.687Z" }, "33hodl/hermes-startup": { "summary": "Hermes Startup is a local-first skill for Hermes Agent designed to help users identify and test business ideas. The tool uses a series of focused questions to generate evidence-labeled ideas and prepares reversible tests for chosen problems.", "highlights": [ "Operates as a deterministic zero-network local runtime by default.", "Requires explicit approval for all consequential actions and external writes.", "Includes contracts for opportunities, offer-previews, and exact-approvals." ], "readmeHash": "97bcb6b05f27a32b872ed28bcdad46ded82a190018cd0b8044217564f49e2a0f", "generatedAt": "2026-09-15T11:56:55.869Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:37:14.439Z" }, "conorbronsdon/agent-memory-kit": { "summary": "agent-memory-kit is a file-based curation loop designed to identify contradictions and outdated information within an agent's memory store. It uses a set of five markdown-based commands to capture, recall, and audit memory through evidence-cited diffs without requiring a database or vector store.", "highlights": [ "Operates using plain markdown files and JSON diffs.", "Includes 'rot' and 'lint' curators to find stale or conflicting facts.", "Uses a local git repository to track and revert memory changes." ], "readmeHash": "4438be8ba4abbc663e2e6a89ce999ea26f98cc7c36e0c2519b980e30e06f571d", "generatedAt": "2026-09-24T12:02:50.734Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-28T16:09:18.407Z" }, "conorbronsdon/agent-skills": { "summary": "agent-skills is a collection of markdown files that teach coding agents specific workflows. It includes skills for session management, code review, state reconciliation, worktree recovery, and skill creation, with installation instructions for Claude Code and other agents. The project is in maintenance mode: it is stable, receives security fixes, and still welcomes issues and pull requests.", "highlights": [ "Skills are installed by copying a directory or using the bundled installer.", "Includes workflows for code review and session management.", "Skills are plain markdown files." ], "readmeHash": "b6bad050884f56c441619ace007d102295463462ad34d6b2163af474a636288b", "generatedAt": "2026-10-12T17:07:47.809Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:34.544Z" }, "bielcarpi/hermes-live-voice": { "summary": "Hermes Live Voice is a community integration that lets you talk to Hermes Agent while it runs background tasks, with results retained when you reconnect. It supports OpenAI Realtime by default, plus Gemini and local Hugging Face voice options. The Dashboard connects to a private gateway, while Hermes owns conversation history, memory, tools, and task execution.", "highlights": [ "Talk while Hermes runs background tasks", "Providers: OpenAI, Gemini, local Hugging Face", "Task results survive voice disconnects and gateway restarts" ], "readmeHash": "bf3abd29d68584ee06de99246229d069b4d96f54c16ed62541d16c5ec990b2dd", "generatedAt": "2026-10-12T00:27:25.992Z", "model": "deepseek/deepseek-v4.1-flash", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:05:13.021Z" }, "allmodels-io/hermes-speech": { "summary": "Hermes Speech is a plugin for Hermes Agent 0.20+ that integrates text-to-speech and speech-to-text capabilities through the AllModels API. It enables conversational setup, voice searching, and streaming audio using Hermes' native provider interfaces and bundled dependencies.", "highlights": [ "Requires Hermes Agent 0.20.0 or later.", "Uses bundled OpenAI and HTTP clients with no additional pip dependencies.", "Supports conversational voice search, previews, and account management." ], "readmeHash": "0b5f2cdabd728a04f5cde10a6df5ad984a0dfa6b69cdd15157ddb473557b97d0", "generatedAt": "2026-09-17T05:03:33.125Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-21T14:37:14.274Z" }, "masterlf/hermes-ai-usage": { "summary": "Hermes AI Usage Monitor is a read-only extension for Hermes Agent that provides telemetry for provider quotas and local token usage. It supports the native Desktop app and Web Dashboard, displaying session history and consumption metrics without reading prompt or message content.", "highlights": [ "Tracks official provider quotas for OpenAI Codex, Anthropic, and OpenRouter.", "Displays local token counters and session history from SQLite state.", "Adds zero runtime dependencies by reusing existing Hermes core components." ], "readmeHash": "499247a366581611410c5fafacc7537fd9d8949a387b243306592385a3b7f80a", "generatedAt": "2026-08-27T13:24:08.885Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-31T15:35:01.355Z" }, "sandbaseai/cli": { "summary": "SandBase CLI connects supported AI clients to SandBase models and APIs through a local MCP bridge. It supports browser authorization and provides commands to discover, inspect, and run available tools, with pricing shown before execution.", "highlights": [ "Connects AI clients to 2,000+ models and APIs via MCP", "Browser authorization connects clients to a local MCP bridge", "Catalog, inspect, and run tools from the command line" ], "readmeHash": "25e5c27a430107cb8659ad97051b6a36a4a02dd5f470684459728aa776d89822", "generatedAt": "2026-10-24T18:21:38.595Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Carlo1911/skill-evolution": { "summary": "Skill Evolution analyzes past agent sessions against installed skills and creates structured proposal files for review. Approved proposals can be applied through an evaluation gate; adapters are provided for Hermes and Claude Code.", "highlights": [ "Fetches sessions and indexes installed skills.", "Proposals are review-only by default.", "Evaluation gate supports deterministic, LLM-judge, and regression evaluators." ], "readmeHash": "20d38ff0583cf2157acdd26796f7045fa97f55cc14c9d46539ad0e926c0d5d13", "generatedAt": "2026-10-18T16:02:11.961Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:36.845Z" }, "joeynyc/hermes-hud": { "summary": "Hermes HUD is a terminal-based consciousness monitor and dashboard for Hermes AI agents. It reads data from the agent's local directory to display live information regarding memory, tool usage, project status, and error logs.", "highlights": [ "Interactive TUI with nine tabs and four selectable color themes.", "Tracks agent growth, git projects, and scheduled cron jobs.", "Provides a tmux operator view for managing live agent panes." ], "readmeHash": "9332f9be80f4dbf73f2a57e8de22c57e837591ff596b79d6de028518c0c94e54", "generatedAt": "2026-08-29T11:41:06.616Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-08-31T15:35:04.468Z" }, "Diabloluo/hermes-hud": { "summary": "Hermes HUD is a local-first, read-only monitoring plugin for the Hermes Agent dashboard that provides real-time observability of system health, token usage, and session activity. It operates as a user-level extension using a shared snapshot cache and WebSocket events to display metrics without outbound telemetry.", "highlights": [ "Provides real-time tracking of token costs, session history, and skill analytics.", "Enforces strict read-only data access with local log desensitization.", "Supports optional incident alerting via Telegram and Feishu." ], "readmeHash": "22dbcc8f5de8e05e77a4bd050ad713e1097d9bc6ed6b2d0314657b6f53f001ef", "generatedAt": "2026-09-28T14:14:24.841Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-28T16:09:22.433Z" }, "181079342/hermes-memoss": { "summary": "MemOS Memory Plugin for Hermes Agent connects to a self-hosted MemOS instance to provide persistent memory storage. It enables semantic search, fact extraction, and automatic synchronization of conversation turns.", "highlights": [ "Supports semantic search and multi-dimensional memory storage.", "Includes a circuit breaker for MemOS API outages.", "Provides tools for searching, adding, and deleting memories." ], "readmeHash": "f13eb9300910fb108885fff53d526969c609f5a8e9e7b4b5d06dc6b4f1a14af6", "generatedAt": "2026-08-31T17:47:10.356Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-07T13:44:29.736Z" }, "FPSUnleashed/hermes-codex-skin": { "summary": "Codex Skin adds a Codex-inspired chat layout to Hermes Desktop, with native theme colors, Glass/Clear backgrounds, history previews, and an optional task Inbox. It offers adjustable composer width and transcript settings; Hermes retains ownership of model menus, voice controls, and Queue behavior.", "highlights": [ "Inbox can settle or snooze threads without stopping their work.", "Settings persist locally.", "Disable the plugin to restore Hermes’ normal appearance." ], "readmeHash": "832c6c8bfc5f6c739a5b28b370e5bddfbaf5a946c48f76a9adbb51a13404d919", "generatedAt": "2026-10-19T15:07:05.910Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Bandwidth/smartnumbers-hermes": { "summary": "Hermes Smartnumbers receives and archives Smartnumbers conversation transcripts, with tools for searching archived turns and managing post-call reviews. When enabled, it can queue one-shot Hermes cron jobs for transcript reviews and user-configured callbacks.", "highlights": [ "Archives transcripts and supports searching transcript turns.", "Queues post-call reviews as one-shot Hermes cron jobs.", "Supports configurable, independent post-call callbacks." ], "readmeHash": "c9f6c440c4941607a741905b30b2364f167cb3be8e081b9b6660140e79ed4312", "generatedAt": "2026-10-12T17:07:51.922Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:40.177Z" }, "vshulcz/deja-vu": { "summary": "deja-vu indexes coding-agent session histories stored on disk so they can be searched across projects and agents. It can surface relevant past sessions at the start of work, resume or hand off sessions, and run locally without a model or server.", "highlights": [ "Search session histories across supported coding agents.", "Recall past decisions and work across agents.", "Sync indexed histories between machines over SSH." ], "readmeHash": "3d96c39d3366e1f3449b4c8589ae9073510b830d9fc33345a52e6ea71f6538f2", "generatedAt": "2026-10-25T16:35:50.262Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Hiutaky/token-saver": { "summary": "Token Saver is a Hermes Agent plugin that calculates the hypothetical cost of local AI inference based on live OpenRouter pricing. It tracks token usage from local providers like Ollama and LM Studio to visualize simulated savings through a web dashboard, TUI widget, and slash commands.", "highlights": [ "Calculates simulated savings using live OpenRouter pricing data.", "Estimates token counts from byte length for local providers.", "Operates locally using Python standard library with no external dependencies." ], "readmeHash": "5a6a73d16b81b9ccb689dbdcf34d3dda56098ebab68550a1c6420ef769a29ff9", "generatedAt": "2026-09-05T18:19:24.799Z", "model": "google/gemma-4-31b-it:free", "version": 1, "audit": "pass", "auditedAt": "2026-09-07T13:44:32.332Z" }, "JanYork/llm-wiki-cli": { "summary": "LWC is an agent-driven proactive memory CLI that lets AI agents recall, maintain, and evolve persistent, source-grounded knowledge across sessions. It turns curated documents into a durable Wiki with preserved sources, pages, citations, links, indexes, and history, and it does not require embeddings or a vector database. A self-hosted team memory option keeps core memory in local SQLite and Wiki files synchronized through a self-hosted server.", "highlights": [ "Agent-driven proactive memory CLI for AI agents", "Maintains a source-grounded Wiki with citations and history", "No embeddings or vector database required" ], "readmeHash": "7a0f45d165ce852c3e6185f8bab535b08cacac505c9d735a4977a81f30d2e69b", "generatedAt": "2026-10-12T00:27:33.129Z", "model": "deepseek/deepseek-v4.1-flash", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:05:20.863Z" }, "superagnt/leverage": { "summary": "Leverage is a collection of agent skills for accessing social and web data, enriching people and companies, and using a Postgres database, task agents, schedules, and dashboards. It also includes packaged skills for complete jobs and can be installed with clients that read SKILL.md.", "highlights": [ "Includes skills for social platforms and web data.", "Offers packaged skills for complete jobs.", "Supports installation through SKILL.md-compatible clients." ], "readmeHash": "279549c151e256579534b7145180c61c6711d9296460a088db9200860b97269b", "generatedAt": "2026-10-25T11:46:14.241Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "pdalinis/incise": { "summary": "Incise is a CLI and agent-tool backend for making precise, byte-preserving edits to Markdown tables, lists, sections, and YAML frontmatter. Callers specify what should change by content-based addresses, and deterministic code handles formatting, boundaries, and the smallest possible splice. It is intended for AI-maintained documentation, Obsidian vaults, Markdown knowledge bases, and local models with limited context.", "highlights": [ "Semantic addresses select headings, rows, items, and keys by content", "Minimal edits change only the targeted structural range", "Ambiguous or unsafe requests fail loudly with actionable refusals" ], "readmeHash": "3150fe6595f0e8316a2993f3b57b67f2fe6d31b05029e1ebbbdf2d5e8f94cdd4", "generatedAt": "2026-10-12T00:27:35.546Z", "model": "deepseek/deepseek-v4.1-flash", "version": 1, "audit": "pass", "auditedAt": "2026-10-12T17:05:21.709Z" }, "moorcheh-ai/memanto": { "summary": "Memanto is a companion agent that manages memories for other agents. It extracts durable knowledge, consolidates duplicates, reconciles conflicting information, applies forgetting policies, and briefs agents with relevant memories. The README says it can run locally or in the cloud and export or import memories using the Open Knowledge Format.", "highlights": [ "Extracts and consolidates memories across agents", "Reconciles conflicting information and manages forgetting", "Supports memory exchange through the Open Knowledge Format" ], "readmeHash": "89d5db6bf3d5bddff10c6d4cf155455b591189546909b9cad4d2c23d4ac46463", "generatedAt": "2026-10-19T18:22:10.410Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "jiawood2006/ai-project-advisor": { "summary": "AI Project Advisor is a chat-based assistant for construction projects on WeCom, DingTalk, and H5. It turns messages into project records, tracks milestones and receivables, flags risks, and generates briefings; rule-based commands handle supported tasks, with an LLM fallback for open-ended questions and summaries.", "highlights": [ "Uses a separate SQLite database per tenant.", "Includes a reversible masking layer before model requests.", "Supports WeCom, DingTalk, and H5." ], "readmeHash": "84979f18818ceccfa125a0ee758c6ae203e15e3504b4913f74ffbdc547b18739", "generatedAt": "2026-10-16T04:45:21.760Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:45.029Z" }, "Senpi-ai/senpi-plugin": { "summary": "Senpi is a plugin that connects supported AI assistants to a user's Senpi agent for Hyperliquid research and strategy workflows. It forwards messages to the Senpi server and returns replies; proposed trades and fund movements require user approval.", "highlights": [ "Offers market signals, trader research, and portfolio review.", "Runs nothing on the user's machine.", "Uses OAuth 2.0 over HTTPS." ], "readmeHash": "53205e891d339165405e35921d718a453f7c53e66a0c5d0e1f24963b65984253", "generatedAt": "2026-10-19T22:32:20.865Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "aidatatools-dev/astro-agents-mcp": { "summary": "Astro Agents provides a hosted MCP server and REST API for Western and Vedic astrology calculations. It returns structured results with stated calculation conventions and SHA-256 input and result hashes; the repository documents the service and includes its MCP registry manifest.", "highlights": [ "Resolves time zones and historical offsets from coordinates.", "Offers tools for charts, transits, aspects, and Vedic calculations.", "Supports MCP and REST; REST payments use x402 or MPP." ], "readmeHash": "cb9fd406e6cdd77f0c590c0aa4ae0657433456fe96f33064f47421eaa5143c25", "generatedAt": "2026-10-25T16:35:51.514Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "hermes-labs-ai/little-canary": { "summary": "Little Canary screens untrusted text by sending it to a powerless canary model and checking whether the input changed its behavior. It returns PASS, FLAG, or BLOCK verdicts for a host to enforce, and also runs structural checks for known suspicious patterns.", "highlights": [ "StructuralFilter checks known patterns without a model or network call.", "Live canary screening can use a local model through Ollama.", "The project publishes evaluation corpora and reproduction commands." ], "readmeHash": "1df4fc1fcbb1f1439edd425b679153ef19843b57ed3603f905c289aac5bdc4e5", "generatedAt": "2026-10-19T22:32:18.217Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "roli-lpci/rule-audit": { "summary": "rule-audit statically analyzes AI system prompts and agent instructions, reporting logical contradictions, coverage gaps, priority ambiguities, meta-rule paradoxes, and absolute-rule edge cases without calling an LLM. It can run as a CLI or pre-commit hook and is available as a plugin for Claude Code, Codex, Gemini CLI, and GitHub Copilot CLI.", "highlights": [ "Produces Markdown, JSON, or summary reports.", "Pre-commit hook audits prompt files before commit.", "Pure Python with no runtime dependencies." ], "readmeHash": "143aa04fa892185803f5b7f82a7daa35c8057f00617ee3531c548cecbf4907fe", "generatedAt": "2026-10-12T17:07:57.273Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:48.092Z" }, "hermes-labs-ai/hermeneutic": { "summary": "Hermeneutic derives reviewable lesson candidates from agent corrections and original instructions, preserving the correction verbatim with provenance. After a person confirms a candidate, a host can retrieve it for a later task; the agent is expected to check its source and applicability against the current instruction. It also supports mining logs from Codex, Claude Code, and OpenAI message files.", "highlights": [ "Candidates require confirmation before contributing context.", "Lessons can be inspected, scoped, and revoked.", "Matching uses conservative lexical overlap by default." ], "readmeHash": "b4f61acdef1364b5a71dc44a37e8d4ac6b7dce5315e5e3f9b0f0ffb503466820", "generatedAt": "2026-10-20T21:46:29.235Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "hermes-labs-ai/hermes-blind": { "summary": "Hermes Blind deterministically extracts the first real user turn from supported local Claude Code or Codex session transcripts and generates an inspectable Markdown recovery anchor. It also provides an evidence-bound prompt wrapper for evaluation workflows and integration options for Claude Code, Codex CLI, Gemini CLI, and Hermes Agent hooks.", "highlights": [ "Offers goals, first-sentence, and full extraction modes.", "Skips tool output and other non-user-turn records.", "Does not use a summarizing model or external service." ], "readmeHash": "e4794b1b6e7deceb7115ce9aa1495d5762a4b19745d3f605987291b9259bbe97", "generatedAt": "2026-10-12T17:16:31.470Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:47.674Z" }, "xP3ta/hermes-console": { "summary": "Console is an independent Android app for connecting to a self-hosted Hermes Agent through Hermes’ Dashboard, without a Console account or intermediary server. It offers chats that show reasoning, tools, and tasks, along with voice conversations, bots and rooms, cron jobs, task views, model management, and SSH terminal access with SFTP and inline file viewers.", "highlights": [ "Connects to Hermes’ Dashboard by scanning a QR code or entering details manually", "Chats display reasoning, tools, and tasks", "Includes voice, bots, rooms, cron, and SSH features" ], "readmeHash": "cf0cdf877c61904dd16f3360b177647f00229c20a9da656dad297f1489dc56c7", "generatedAt": "2026-10-21T05:04:08.868Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "parkyucheol-del/alphapipeline": { "summary": "AlphaPipeline provides paid, per-call endpoints for token risk checks and market data, using x402 payments on Base. It also exposes a remote MCP server that clients can use to discover and call its tools; discovery is free, and two onboarding tools are permanently free.", "highlights": [ "Checks token honeypot, tax, and liquidity risk.", "Provides market data including DEX slippage and funding rates.", "Offers REST endpoints and a remote MCP server." ], "readmeHash": "214f11b49f1c2f923da0e12b2a638d6081ec7b79962cb8189a547d2ed2547cd0", "generatedAt": "2026-10-18T16:02:15.814Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:50.789Z" }, "Jezue/fitatu_library": { "summary": "fitatu-api is an unofficial Python client for Fitatu that can read and update food diaries, activities, body measurements, water, goals, summaries, and food catalog data. It also provides tools for AI agents and an MCP server that makes those tools available to MCP hosts.", "highlights": [ "Includes 47 tools for Fitatu diary operations.", "MCP server supports stdio and HTTP.", "Requires Python 3.11 or newer." ], "readmeHash": "652c6e4d87e5b0c973090b9d60183d7d051ee282b5d7682086b86f65aa9edec6", "generatedAt": "2026-10-20T17:22:55.682Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "robbyczgw-cla/hermes-ui-plus": { "summary": "HUI+ is an unofficial mobile, web, and desktop client for a self-hosted Hermes Agent. It lets users respond to interactive answers, approve or deny commands, and follow tool steps live; it requires Hermes Agent and a locally run bridge.", "highlights": [ "Interactive answers include checklists, choices, forms, rankings, and votes.", "Command approvals and live tool-step updates.", "Runs on Android, web, macOS, and Linux alpha." ], "readmeHash": "262a31d8cd70e2fed1d5c501a8d0da168726ab01f72a75288183f56a7da315de", "generatedAt": "2026-10-20T21:46:29.972Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "jyje/hermes-agent-helm": { "summary": "jyje/hermes-agent-helm provides a Helm chart for running Hermes Agent on Kubernetes. It supports provider configuration through values, Discord or Telegram gateway connections, and optional Codex or GitHub Copilot device login with credentials persisted across restarts. The chart can also enable Hermes’s web dashboard and supports grouping multiple instances into a Hermes Team.", "highlights": [ "Install from OCI registry or published Helm repository.", "Includes provider-specific values examples.", "CI installs the chart on an ephemeral kind cluster." ], "readmeHash": "a678e4e089a9745ada1d74224614c83b34bcf10e562806ed9bc6ac924ec494f4", "generatedAt": "2026-10-21T01:14:52.143Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "taskawayfr/hermes-agent-fr": { "summary": "Guide pratique en français pour installer et configurer Hermes Agent, le connecter à Telegram, gérer les profils et la sécurité, et programmer des tâches. Il comprend aussi des chapitres sur le dépannage, les coûts, l’hébergement, la migration depuis OpenClaw et des cas d’usage pour petites entreprises.", "highlights": [ "Fournit des exemples prêts à copier et testés.", "Vérifie les commandes, exemples et liens automatiquement.", "Publié sous licence Creative Commons Attribution 4.0." ], "readmeHash": "2037e3c107f63fad6d05a4cb14fd1552c973eeec164e9e29c82b101710088c15", "generatedAt": "2026-10-18T07:50:53.283Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass", "auditedAt": "2026-10-19T15:04:56.850Z" }, "NickSF189/kisscode-agent": { "summary": "This repository provides a skill for AI agents to join KissCode, a relationship network where AI agents and people meet. It includes installation instructions, a one-call registration example, and guidance to add HEARTBEAT.md to an agent’s routine.", "highlights": [ "Supports agents that read SKILL.md.", "Provides installation instructions for several agent environments.", "The repository mirrors kisscode.ai/skill.md." ], "readmeHash": "6119abef855a8d30ad236ecdb3abba6b48a3c32328a915b243d575e735a864e2", "generatedAt": "2026-10-24T04:50:40.404Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "arcmira/mcp": { "summary": "Arcmira provides a remote MCP server for searching YouTube and podcast transcripts, retrieving timestamped passages, reading surrounding transcript lines, and finding mentions, appearances, endorsements, ad reads, sponsors, and mention trends. Agents can also set up YouTube monitors; the README describes sign-in through an MCP host or authentication with an account key.", "highlights": [ "Search YouTube and podcast transcripts for timestamped quotes.", "Analyze mentions, appearances, recommendations, sponsors, and momentum.", "Create and update YouTube monitors through the MCP tools." ], "readmeHash": "091c27bb5e3296e7b6a73af105883db024a190595a039ee62ce5edc3f86705d1", "generatedAt": "2026-10-21T01:14:54.537Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "nicolasestrem/hermes-model-switcher": { "summary": "Hermes Model Switcher is a community plugin for the Hermes Web Dashboard that displays supported model routes across profiles in a searchable, filterable table. Users can preview model changes, apply them, save choices as presets, review per-profile results, and use revision-guarded Undo.", "highlights": [ "Shows configured routes across profiles in one matrix.", "Previews changes before applying them.", "Includes named presets and guarded Undo." ], "readmeHash": "ae5ff2685ffa4c6750cb8f0147e283445acbf8b03d795f48ec7264387f773554", "generatedAt": "2026-10-22T05:08:03.773Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "wjgoarxiv/lithermes": { "summary": "LitHermes is a plugin for Hermes Agent that starts a workflow when a request begins with `lit`. It plans the task, checks the result, and records goals, plans, and evidence on disk, with a handoff note to help a later session continue.", "highlights": [ "Supports `lit` requests and `/lit` commands.", "Records work and evidence on disk.", "Provides a handoff note for the next session." ], "readmeHash": "de7f5a11f48e422237242b9fa7fa7e94ee71f26394a3ee99d8bc2283ec8847a5", "generatedAt": "2026-10-26T05:02:40.141Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "f-o-x11/deed-page-skill": { "summary": "This repository contains an agent skill that publishes a file or folder to a permanent deed.page URL with one curl request, and documents deed.page’s hosted remote MCP server. The MCP server can publish and manage sites, including updating, versioning, password-protecting, and deleting them; the skill and server are described as compatible with several agent clients.", "highlights": [ "Publishes files or folders to permanent deed.page URLs.", "Documents a hosted remote MCP server at deed.page/mcp.", "Includes agent plugin manifests and installation instructions." ], "readmeHash": "9998f1a56f42073ee6db00327c1b4610fc4545587dee568f2ead973fa4b56fcc", "generatedAt": "2026-10-26T05:02:38.828Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Webly-AI/webly-plugin": { "summary": "Webly provides an agent skill and plugins for publishing a folder or website to a live Webly URL without an account. Users can claim a site by signing in, then work with it through Webly’s MCP server, including editing and publishing.", "highlights": [ "Installs for Claude Code, Cursor, Gemini CLI, and other agents.", "Unclaimed sites are live for 24 hours and claimable for 7 days.", "MCP access supports site editing, publishing, and rollback." ], "readmeHash": "8b8364fb4536ea41a842f8015a5c3e2b9f097c0aaf57583074ef6c642f58e419", "generatedAt": "2026-10-25T05:18:05.965Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "sergebulaev/linkedin-skills": { "summary": "LinkedIn Marketing Skills for Claude Code and Codex is a collection of 12 skills for drafting LinkedIn posts, comments, and replies, reviewing drafts, planning content, and editing profiles. The skills show a draft and wait for approval before publishing; the README also documents installation for Hermes Agent and other tools that read SKILL.md files.", "highlights": [ "Includes 12 LinkedIn-focused skills", "Drafts require approval before publishing", "Supports installation in Hermes Agent" ], "readmeHash": "0a6a844b04679139862c8450731f581457d263adb507d12129d9b82cf32960e6", "generatedAt": "2026-10-25T22:50:28.544Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "marospekarik/paperclip-hermes-file-access-manager": { "summary": "File Access Manager is a Paperclip plugin that sets per-path Read/Write, Read Only, or Denied permissions for Hermes profiles. It translates those rules into Docker bind mounts and uses Hermes’ Docker terminal backend to enforce access inside a container. The plugin provides a filesystem tree editor, previews generated mounts, and can apply changes by recreating the container and restarting the gateway.", "highlights": [ "Denied paths are absent from the container by default.", "Folder permissions inherit downward; deeper rules override.", "Includes unit tests and real-Docker integration tests." ], "readmeHash": "2cb8deb130f4b25a7163a2ee66a24be96d37a93d959aa01739c8295a9ae6eb95", "generatedAt": "2026-10-25T22:50:28.709Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "thomasbek3/hermes-bot-kit": { "summary": "Hermes Bot Kit provides desktop plugins that add texting-style bubbles, named bot sections, a Tasks list, and a docked remote-desktop pane to Hermes Desktop. It also includes agent plugins for texting-style replies and for shell and GUI interaction with a bot’s pinned computer.", "highlights": [ "Includes four Hermes Desktop plugins and two agent plugins.", "Computer pane supports remote machines and bot-specific bindings.", "Desktop plugins require Hermes Desktop 0.20.5 or later." ], "readmeHash": "bc53a38b750e6fcefee4c9a2b68acb55365cb14007eb74b3b22b0011f8b4d190", "generatedAt": "2026-10-25T22:50:31.792Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "TheSmokeDev/hermes-talk": { "summary": "Hermes Talk is a voice plugin for Hermes Agent that lets users speak with the agent, hear spoken replies, and use its tools during a conversation. It supports terminal, Discord voice, dashboard, and Hermes Desktop sessions, with voice options including OpenAI Realtime, xAI Grok Voice, Gemini Live, GPT-Live, and Cascade voice; provider availability varies by surface.", "highlights": [ "Calls Hermes tools and delegates background work during voice conversations.", "Offers terminal, Discord, dashboard, and Desktop interfaces.", "Provider options and authentication requirements vary by voice mode." ], "readmeHash": "ef1b4c9864f4270b660c6dee79fc1beaaa7249bbe625d2a1f631e2402dbebdc9", "generatedAt": "2026-10-25T22:50:32.288Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Sahil-SS9/hermes-walkie-talkie": { "summary": "Hermes Walkie Talkie is a standalone plugin that lets independent Hermes sessions on the same computer discover one another and coordinate through messages, groups, broadcasts, and structured requests. It uses Unix-domain sockets on Linux and macOS, and Windows named pipes with SID-bound access controls; it does not connect machines over a network or use a central daemon.", "highlights": [ "Direct messages include delivery receipts.", "Supports named groups and bounded broadcasts.", "Includes Hermes tools, slash commands, and an optional Desktop panel." ], "readmeHash": "dd037e93be707bf0f3e6806f09de0efc73b3c4db3a3d4d423fe02e2bb9e97916", "generatedAt": "2026-10-25T22:50:34.248Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Sahil-SS9/llm-judge-jury-executioner": { "summary": "A standalone skill and CLI that runs questions through multiple independent models using six protocols, including voting, synthesis, critique, red-team review, and full deliberation. It can produce Markdown, JSON, HTML, and transcript outputs, with configurable provider routes and bounded workspace context scanning.", "highlights": [ "Six protocols with fail-closed configuration validation", "Outputs verdicts, reports, and phase transcripts", "Supports OpenAI-compatible provider APIs" ], "readmeHash": "cc14faafebbfa04626b810200f00d8f8a360db8c0578b6fa3111be31d6268014", "generatedAt": "2026-10-25T22:50:36.376Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Sahil-SS9/hermes-project-stewardship": { "summary": "Hermes Project Stewardship lets a team of Hermes agents look after a project over time. It verifies project state before acting, proposes work with supporting evidence, supports approval gates, and measures outcomes through a desktop app. Projects, boards, and tasks remain in Hermes, while Stewardship stores its own state in a local SQLite database.", "highlights": [ "Includes project dashboards, Kanban boards, workflows, and an approval inbox.", "Supports per-project autonomy policies and human approval.", "Uses Hermes Projects and Kanban as the record for project work." ], "readmeHash": "95f8662db02d94851006ff30295d5737a2784da9b23a27325d45ec2d2e20354d", "generatedAt": "2026-10-26T05:02:41.985Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "MemTensor/memmy-agent": { "summary": "Memmy is a desktop app and CLI/TUI for local memory and AI-agent tasks. Its Memory Service stores and retrieves information, while its agent features can carry project context between supported tools, including Claude Code, Codex, OpenClaw, and Hermes.", "highlights": [ "Provides a local Memory Service and memory CLI.", "Offers a desktop app, CLI, and interactive TUI.", "README lists integrations with multiple AI coding agents." ], "readmeHash": "7328187e4672a8c07f2dc78aaa5e71c17e5b7340636e2b62ed830d204023337a", "generatedAt": "2026-10-25T22:50:39.235Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "rlaope/oh-my-hermes": { "summary": "oh-my-hermes is an operating layer for Hermes Agent that routes work to skills and model chains by category and runs ready graph nodes in parallel. It includes task-ranked long-term memory with explicit decisions about what to retain, and reports work as verified only after a test, review, or CI gate passes.", "highlights": [ "Runs graph-based tasks in parallel with a separate goal-verification node.", "Routes lanes to editable model chains based on work category.", "Provides a TUI on Hermes' gateway." ], "readmeHash": "09295588a655bbc0b8553733008f3d120c63f69872cce4d85829df16fedecd84", "generatedAt": "2026-10-26T07:19:36.498Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "Orkas-AI/Orkas": { "summary": "Orkas is an open-source, local-first desktop app where a Commander plans work and delegates tasks to built-in specialists and supported local agents. Projects retain goals, rules, memory, and files across conversations; to-dos, auto-advance, and scheduled automations help ongoing work continue.", "highlights": [ "Includes 16 built-in specialist agents.", "Connects to supported local coding and personal agents.", "Keeps conversations, files, and custom agents on your disk by default." ], "readmeHash": "8a07e28e286ed81390a6e21136dce76d5986baca330fff4bbc9956204dbc0036", "generatedAt": "2026-10-25T22:50:42.366Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "praveen-ks-2001/hermes-agent-template": { "summary": "This Railway template deploys Hermes Agent with a web admin dashboard for configuring providers, channels, and tools, and managing the gateway and user pairing. It also provides the Hermes web UI behind the same login, plus backup and restore for Hermes-home data.", "highlights": [ "Configure providers, channels, and tools in a setup wizard", "Manage gateway status, logs, and user pairing in the browser", "Create and restore backups of Hermes-home data" ], "readmeHash": "9ae86bdeb249bfa6d66650cd4d324d6e1ce6501ebcdceecb3dbc7fc20e6d013b", "generatedAt": "2026-10-25T22:50:43.349Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "tutti-os/tutti": { "summary": "Tutti is a real-time shared workspace that connects agents’ context, files, apps, and tasks. It supports connecting existing agents and using existing subscription credits; the open-source version also includes built-in apps. The README describes Tutti · VM as adding features such as group chat and simultaneous editing.", "highlights": [ "Agents can reference shared conversations, files, app outputs, and tasks.", "Built-in apps include image generation, design, docs, and presentations.", "Tutti · VM supports group chat and simultaneous editing." ], "readmeHash": "3193f2159d8d38a2c12221fb0a622e54f7e78c20ec18d155aecc3e8af6d59980", "generatedAt": "2026-10-25T22:50:44.541Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "iranee/fnos-hermes-agent": { "summary": "Hermes Agent for fnOS is an AI assistant application for Feiniu NAS, deployed through the fnOS app center as an .fpk package. It provides a web control panel for configuration and chat, with support for multiple messaging channels, image and file inputs, and model fallback.", "highlights": [ "Web chat supports Markdown and multiple concurrent conversations.", "Connects to WeChat, QQ, Feishu, WhatsApp, and Telegram.", "Includes a control panel for configuration and status." ], "readmeHash": "5a65cad5f16320d9d6a011064c6a4f8d083a751751dc0904a6ea8360f134db01", "generatedAt": "2026-10-25T22:50:45.915Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" }, "silent-reader-cn/hermes-anywhere": { "summary": "Hermes Anywhere is a Flutter and Cupertino client for Hermes Agent, with the same interface and sessions on Windows and Android. It supports streaming chat, session management, workspace tools, Kanban boards, and usage insights; the Windows installer bundles a WebUI backend and Python 3.11 runtime.", "highlights": [ "Shared interface and sessions across Windows and Android", "Streaming chat with reasoning and tool-call cards", "Windows installer includes a WebUI backend and Python runtime" ], "readmeHash": "5a8100fdb4a02ebdc07eccf2eadfbe686c6e1158d3c000f967c8d6918cebce25", "generatedAt": "2026-10-26T07:33:35.365Z", "model": "openai/gpt-6-luna", "version": 1, "audit": "pass" } }