The most popular skills and skill registries in the Hermes ecosystem, ranked by GitHub stars. Skills are reusable capabilities that follow the agentskills.io open standard — from cybersecurity to diagram generation to creative ideation.
This page ranks the whole category by stars. For picks by what you're trying to do — with install commands, compatibility notes, and a verdict per skill — see the Hermes Skills Hub.
This framework implements a subagent-driven development methodology designed for autonomous task execution. It enforces test-driven development and YAGNI principles while integrating as a plugin for various coding harnesses and CLIs.
A local-first design and prototyping system that generates web, mobile, and motion graphic artifacts in formats like HTML, PDF, and MP4. It features native agent runtime support through integration with the DeepSeek Harness.
This repository provides 818 structured skills across 34 security domains, mapped to frameworks including MITRE ATT&CK, ATLAS, and NIST CSF 2.0. The library is compatible with over 26 AI platforms for standardized security operations.
This skill converts codebases, infrastructure-as-code, and schemas into draw.io diagrams using natural language descriptions. It supports incremental synchronization to preserve manual layouts and includes both a CLI and an optional MCP server.
Acting as a trust layer for agent-to-agent commerce, this project provides a master router and connector skills for delegated permissions and payments. It incorporates a verification and dispute resolution layer to manage natural-language mandates and escrow.
This tool audits and rewrites content to eliminate detectable AI patterns using a two-pass detection system. It supports rewrite, detect-only, and edit-in-place modes with optional voice profiles for different tones.
This agentic evolver automates skill deduplication and quality improvement across sessions for Hermes and other OpenAI-compatible APIs. It includes a bilingual dashboard for inspecting session traces and monitoring collective skill evolution.
A cross-platform library featuring 51 expert frameworks and 14 guided metaskills compatible with Claude Code and agent-plugins.org specifications. It maintains session state locally within the project's documentation folder.
This collection contains 1,169 professional and life skills distributed as plain markdown files under the MIT license. The skills are designed to run natively in Hermes Agent and Claude Code with ready-to-paste exports for other assistants.
This tool enables agents to perform video searches, browse channels, and extract transcripts or playlist data. It is compatible with Hermes Agent, OpenClaw, and various AI-powered IDEs, requiring a TranscriptAPI key for operation.
A meta-skill plugin that automatically detects and proposes reusable skills by watching active workflows. It generates both SKILL.md documentation and Python scaffolds, allowing users to manage detected patterns via slash commands.
This project teaches Hermes Agent to write its own analytical prompts and custom thinking instructions for specific tasks. It utilizes pre-built analytical lenses and generates reports that highlight potential analysis blind spots.
This project enables type-safe agentskills.io support for Pydantic AI with remote registry integration via Git and S3. It provides tools for reading bundled resources and executing untrusted scripts within a sandboxed environment to isolate them from the host.
This skill set transforms rough concepts into structured design documents and implementation handoffs for Hermes Agent. It features Lite and Full modes for different project complexities and includes a spec review gate that uses PASS/FAIL evaluation models.
This literate programming skill allows agents to weave prose, diagrams, and source code into .lit.md files. It includes a tangler to extract runnable source files and automatically synchronizes edits from those files back into the original documentation.
Maestro is a local-first, Git-based harness for long-running agents that operates without a background service. It utilizes a three-layer system of kernels, plugins, and Markdown recipes to implement the SLP protocol for supervised team coordination.
This plugin facilitates cross-instance shared memory by capturing high-value decisions and outcomes in markdown files. It establishes a pipeline for fine-tuning and transitioning to replacement models based on recorded work history.
This community resource provides installation guides and skills in the SKILL.md format specifically for Portuguese-speaking users. It supports the agentskills.io standard to facilitate localized Hermes Agent integrations.
Following the Agent Skills specification, this project enables agent interaction with Oracle networks and smart contracts. It supports Chainlink services including CCIP, Data Feeds, VRF, and Data Streams via the Vercel open skills ecosystem.
These official skills provide API integration and prompting guidelines for FLUX image and video generation. The suite includes eight specialist skills for FLUX 3 video tasks and supports polling and webhooks for BFL API interactions.
This tool converts repositories, documentation, and developer discourse into agentskills.io-compliant skills. It features automated workflows for briefing, testing, and exporting skills while linking instructions to specific file lines or URLs.
Skilldock.io is a registry for reusable AI skills that supports recursive dependency resolution during installation. It provides a Python client and CLI, incorporating commerce features through TON payments.
Nika defines AI workflows as code using a YAML-based language with four verbs and a Rust binary. This local-first tool includes a CLI for checking and tracing task execution across OpenAI and Codex models.
This skill enables autonomous execution of complex, multi-step tasks with real-time progress and decision tracking. It includes built-in checkpoints and rollback procedures to ensure safe recovery during long-running operations.
This skill enables Hermes Agent to manage files, calendars, tasks, and contacts through standard WebDAV and CalDAV protocols. It ensures secure communication via HTTPS and includes a guided setup script for credential validation.
This progressive MCP client utilizes a three-layer lazy loading architecture to minimize token usage during agent interactions. It features a socket-based daemon for persistent connections and supports incremental discovery of server metadata and tool schemas.
Serving as a community hub for skill discovery and one-click installation, this project implements the Agentic Resource Discovery (ARD) v0.9 specification. it operates autonomously on NVIDIA infrastructure and provides federated search across GitHub and Hugging Face.
This project implements a five-layer retrieval system that uses hard triggers and semantic embeddings to narrow skill lists before LLM calls. By filtering skills through RRF fusion and FTS5, it reduces token usage and supports graceful degradation if dependencies are missing.
Designed as an ADHD-friendly execution system, this project provides agent-driven scheduling and habit tracking via an MCP connection. It stores all data in a local SQLite file and includes both a React web application and an Expo-based Android app.
This read-only skill performs adversarial bug-hunting by deploying three parallel subagents to attack code from different angles. It combines deterministic scans using tools like Semgrep and Bandit with verified findings produced in a hardened sandbox.
This plugin rewrites and scores prompts across multiple surfaces like Discord and Telegram to reduce token usage by 20–80%. It supports 97 languages and evaluates prompt quality across five specific dimensions.
Inspired by Claude Code, this skill uses a swarm of three sub-agents to improve code hygiene, clarity, and correctness across five programming languages. It employs a tiered application process that requires an approval gate before applying changes.
This project introduces systematic investigation protocols and multi-story task decomposition with mandatory evidence-based verification gates. It utilizes a Python goal engine to manage persistent task states and prevent early-stop errors during debugging.
This C# and .NET implementation of the agentskills.io pattern enables tool-based skill loading via Microsoft.Extensions.AI. It supports progressive disclosure through SKILL.md files and automates the registration of common skill directories.
This turnkey laboratory equips Hermes Agent with over 2,000 skills and 131 tools for network forensics and exploitation. It integrates 28 cybersecurity frameworks and standards to support research into CVE-driven exploits spanning from 2015 to 2026.
These production-tested skills provide session management and worktree recovery to maintain memory between agent conversations. The toolkit also orchestrates parallel subagents for architectural reviews and detects state drift in development environments.
This local-first dashboard provides real-time observability for Hermes Agent across 11 monitoring tabs, including token costs and error tracking. It enforces a strict read-only boundary with zero outbound telemetry to ensure data privacy during session and skill monitoring.
This skill automates the creation of reproducible demo GIFs for repositories using vhs or Playwright recording scripts. It handles the full pipeline of rendering, compressing files to under 8 MB, and embedding the resulting markdown into README files.
This tool generates portable agent skills from plain language and validates them against runtime-specific profiles to prevent specification drift. It monitors official frontmatter specs weekly and uses CI-friendly exit codes to report inconsistencies.
This CLI and MCP server allows agents to search a database of over 13,870 MCP servers and 4,384 skills. It provides direct installation commands for discovered tools and integrates with MCP-compatible clients like Hermes and Cursor.
This host-agnostic framework creates a self-improvement loop by analyzing session history to identify and propose skill coverage gaps. It gates all updates behind deterministic and LLM-judge evaluators and operates locally using the Python standard library.
These skills enable deterministic path routing and Python interpreter selection across Windows, Git Bash, and WSL environments. The tool ensures environment-specific path syntax is used and verifies project-compatible runtimes without requiring new installations.
This skill automates USDC payments for HTTP 402 responses using protocols such as Pay Link and x402. It is designed for compatibility with the agentskills.io standard and works across OpenClaw and Hermes Agent environments.
This collection provides specialized utilities for Chinese language processing, including Douyin video transcription and rule-based removal of AI writing patterns. It also features macOS-based OCR for structured data extraction and an e-commerce material studio.
This project provides a functional quality seal by testing agent skills for structure, functionality, and security within isolated Docker containers. It issues cryptographically verifiable Ed25519 certificates to confirm that a skill operates as intended.
This plugin facilitates a decentralized support ecosystem where users earn USDC on the Base network by answering bounties over x402. It allows for listing bounties and creating boards without requiring accounts or API keys, using dedicated wallet keys for transactions.
This integration exposes 191 tools and five production-grade workflow skills through a hosted MCP server. It enables Hermes Agent to access structured web data across 31 different platforms.
This skill generates multi-scale Chinese news reports ranging from global events to street-level updates. It utilizes semantic memory to integrate user interests and applies quality gates to produce structured analysis and action suggestions.