mem0ai / mem0
Mem0 is a memory layer for AI agents with separate memory states for users, sessions, and agents. It is an official Hermes Agent memory provider, available as an OSS SDK, a self-hosted server, or managed cloud.
Best Memory Providers for Hermes Agent
The top community-built memory and context management tools for Hermes Agent, ranked by GitHub stars. Memory is what makes Hermes self-improving — these tools extend its built-in recall with long-term semantic memory, graph-based retrieval, and cross-session persistence.
For the architecture behind these tools, read The Hermes Agent Memory Guidebook.
#
project
stars
01
mem0ai / mem0
The Memory Layer for AI Agents - Drop-in memory infrastructure for AI agents and apps. Context that persists. Built for production.
★ 67.1K
02
vectorize-io / hindsight official
Hindsight: Agent Memory That Learns
★ 48.8K
03
volcengine / OpenViking
Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
★ 39.8K
04
topoteretes / cognee
Cognee is the open-source AI memory platform for agents. Give your AI agents persistent long-term memory with small models for free
★ 32.1K
05
supermemoryai / supermemory
Memory and context engine + app that is extremely fast, scalable, and can be run fully locally. The Memory API for the AI era.
★ 31.2K
06
garrytan / gbrain
Garry's Opinionated OpenClaw/Hermes Agent Brain
★ 30.8K
07
plastic-labs / honcho official
Memory library for building stateful agents
★ 7.6K
08
campfirein / byterover-cli
ByteRover CLI (brv) - The portable memory layer for autonomous coding agents (formerly Cipher)
★ 5.0K
09
mnemosyne-oss / mnemosyne
Zero-cloud AI memory that works everywhere. SQLite-backed. One pure-Python dependency.
★ 3.4K
10
moorcheh-ai / memanto
Memory that AI Agents Love!
★ 2.3K
11
ClaudioDrews / memory-os
A 7-layer memory operating system for Hermes Agent — persistent memory with Qdrant, structured facts, fabric recall, auto-curated wiki, and
★ 1.4K
12
greyhaven-ai / autocontext
a recursive self-improving harness designed to help your agents (and future iterations of those agents) succeed on any task
★ 1.3K
13
stephenschoettler / hermes-lcm
Lossless Context Management plugin for Hermes Agent — DAG-based context engine that never loses a message
★ 1.3K
14
vshulcz / deja-vu
Memory for Claude Code, Codex, Cursor and 38 more coding agents, built from the session history already on your disk. Local search, MCP and
★ 1.2K
15
elkimek / honcho-self-hosted
Self-host Honcho memory layer for Hermes Agent — OpenRouter + Venice, no code changes
★ 381
16
Signet-AI / signetai
Sync and store memories, shared identity files (AGENTS.md, CLAUDE.md), session transcripts, institutional knowledge, and secrets between all
★ 305
17
plur-ai / plur
Shared memory for AI agents
★ 300
18
yoloshii / ClawMem
On-device memory layer for AI agents. Claude Code, OpenClaw and Hermes. Hooks + MCP server + hybrid RAG search.
★ 210
19
Sibyl-Labs / Sibyl-Memory
Durable, file-based long-term memory for AI agents. Five-package plugin family: SDK, CLI, MCP server, Hermes adapter, and a LangGraph BaseSt
★ 172
20
EfficientContext / ContextPilot
Accelerating Long Context LLM Inference with Accuracy-Preserving Context Optimization in SGLang, vLLM, llama.cpp, OpenClaw, RAG, and Agentic
★ 141
21
yantrikos / yantrikdb-hermes-plugin
YantrikDB memory provider for NousResearch/hermes-agent — self-maintaining memory with benchmarked recall, self-tuning ranking, contradictio
★ 96
22
JanYork / llm-wiki-cli
Agent-driven proactive memory CLI for AI agents — autonomously recall, maintain, and evolve persistent, source-grounded knowledge across ses
★ 66
23
ccf / agentcairn
Long-term, cross-project memory for AI coding agents. Your own Obsidian vault as the source of truth. Daemonless and without opaque database
★ 65
24
amanning3390 / flowstate-qmd
FlowState QMD — Anticipatory Memory for AI Agents. Hermes 2026 Hackathon Entry.
★ 51
25
yepyhun / Brainstack
Memory kernel stack for Hermes agent
★ 48
26
generalbusiness-ai / keep
Reflective memory for agents
★ 40
27
aliasocracy / hermes-tool-slimmer
Reduce Hermes Agent tool-schema overhead with keyword selection and Tool Search support
★ 39
28
conorbronsdon / agent-context-os
Portable, evolving context and workflow layer that works across Claude Code, Codex, Hermes Agent, OpenClaw, Cursor, Devin and other compatib
★ 29
29
Ladybug-Memory / hermes-memory-plugin
Ladybug memory plugin for Hermes Agent
★ 24
30
stepanov1975 / hermes-local-knowledge
Hermes Agent plugin for local knowledge search and artifact routing—find the right skill, script, runbook, cron job, MCP server, or document
★ 11
31
MukundaKatta / hermes-agentmemory
Pull-model episodic memory plugin for Hermes Agent. Real deletes, audit trace, BYO Claude. MIT.
★ 10
32
hermes-labs-ai / hermes-blind
Recovers the original goal of a long Claude Code or Codex session from its first user turn, so you can restate it before continuing — plus a
★ 10
33
piprail / piprail
x402 (HTTP 402 Payment Required) SDK + MCP server: let any API charge for itself and any AI agent pay for itself, USDC & stablecoins across
★ 9
34
yonro / hermes-xmemo-plugin
XMemo memory provider for Hermes Agent — durable cross-session AI memory with semantic search
★ 8
35
futurebrowser / hermes-exabase-plugin
Exabase memory plugin for Hermes Agent.
★ 5
36
sourcevault-ai / sourcevault-code-tools
Hermes Agent plugin — private local code memory for your repos (semantic search, file reads, grounded Q&A)
★ 5
37
visorcraft / MongrelDB-Hermes
AI-native long-term memory for Hermes Agent with AES-256-GCM encryption at rest, dense HNSW ANN, sparse and exact-text recall, metadata/rang
★ 5
38
penfieldlabs / hermes-penfield
🧠 Penfield memory provider for Hermes agent
★ 4
39
conorbronsdon / agent-memory-kit
The curation pass for agent memory: a thin, file-based loop that finds rot and contradictions and proposes evidence-cited diffs you review.
★ 2
breakdown
vectorize-io / hindsight
Hindsight provides long-term agent memory through retain, recall, and reflect operations. It offers Python, Node.js/TypeScript, and Go clients, with support for multiple LLM providers, including local options.
volcengine / OpenViking
OpenViking is an open-source context database that organizes knowledge, memories, and skills in one filesystem, with L0 abstracts, L1 overviews, and L2 details. It is an official Hermes Agent memory provider using viking:// URIs; sessions can be archived and memories extracted as Markdown.
topoteretes / cognee
Cognee builds connected memory from documents, code, and conversations for persistent use across agent sessions. It also supports local text processing without an LLM and provides a CLI for remembering and recalling text.
supermemoryai / supermemory
Supermemory provides memory, RAG, profiles, and connectors through one API, with processing for documents, images, video, and code. It is an official Hermes Agent memory provider and offers plugins for several AI assistants, including Hermes.
garrytan / gbrain
Gbrain is an opinionated brain for OpenClaw and Hermes Agent that stores sourced facts and supports corrections and withdrawal. It connects to existing agents and can share memory across them, with optional semantic search, synthesis, and background enrichment.
plastic-labs / honcho
Honcho is a memory library for building stateful agents: it stores messages and events on sessions and reasons in the background to update peer representations. It is the upstream library used by the elkimek/honcho-self-hosted Hermes wrapper, and is available as a managed service, local CLI, or self-hosted FastAPI server.
campfirein / byterover-cli
ByteRover CLI treats memory as a Git repo, managing a context tree with Git-like commit, branch, and merge operations over plain Markdown. It is an official Hermes Agent memory provider and includes an interactive TUI and REPL, plus support for 20 LLM providers and 24 built-in agent tools.
mnemosyne-oss / mnemosyne
Mnemosyne is a SQLite-backed memory layer for Hermes agents, with one pure-Python dependency. It integrates through MCP, native providers, or a Python SDK; BEAM results are reported for v3.0.0, while LongMemEval is pending.
moorcheh-ai / memanto
Memanto consolidates memories and reconciles conflicting information, then briefs agents with relevant memories before they act. It supports importing and exporting through the Open Knowledge Format.
ClaudioDrews / memory-os
Memory OS combines seven memory layers, including vector search and structured facts, with an auto-curated wiki. Its local infrastructure uses Docker, Qdrant, Redis, and SQLite, and its Ground Truth hierarchy is designed to prioritize injected context.
greyhaven-ai / autocontext
Autocontext runs agent tasks against evaluations, keeps useful lessons, and discards dead ends to help agents succeed on complex tasks. It also produces traces, reports, and playbooks.
stephenschoettler / hermes-lcm
This Hermes Agent plugin uses a SQLite store and hierarchical DAG to persist messages, with the goal of never losing a message. Its lcm_grep and lcm_recall tools help recover details from past context.
vshulcz / deja-vu
Deja-vu searches session histories across 41 coding agents, making past work available for recall. Recall can appear at session start or after command failures, and sessions can be synced between machines over SSH.
elkimek / honcho-self-hosted
This project lets Hermes Agent use a self-hosted Honcho backend for private memory that persists across sessions. It supports OpenAI-compatible providers and local inference servers, with an automated setup script for Docker, PostgreSQL, and Redis.
Signet-AI / signetai
Signet provides a local-first identity, memory, and secrets layer with portable state across several agent harnesses. It can import files, webpages, and agent conversations, and includes a dashboard for browsing memories and their sources.
plur-ai / plur
Plur stores shared agent memories as plain-text YAML engrams on disk. It offers local search after a one-time model download, with keyword fallback; its README reports 95.6% R@5 on LongMemEval-S (N=500).
yoloshii / ClawMem
ClawMem is an on-device context and memory engine that combines BM25, vector search, rank fusion, and cross-encoder reranking. It can generate session handoffs and capture decisions and preferences, with integrations through hooks, MCP, and native agent plugins.
Sibyl-Labs / Sibyl-Memory
Sibyl-Memory is a file-based long-term memory system with a SQLite-backed, five-tier hierarchical schema and FTS5 search. Its five-package family includes an SDK, CLI, MCP server, Hermes adapter, and LangGraph BaseStore, without a vector database or embedding model.
EfficientContext / ContextPilot
ContextPilot optimizes context ordering for prefix-cache sharing, including text and multimodal context. Its context index supports reordering, deduplication, and cache-aware scheduling, and it integrates with Hermes Agent, vLLM, SGLang, llama.cpp, and OpenClaw.
yantrikos / yantrikdb-hermes-plugin
This self-maintaining Hermes memory plugin uses an in-process backend, so it does not require a separate server. Recall results include a why_retrieved explanation, and the plugin supports canonicalization and contradiction tracking.
JanYork / llm-wiki-cli
This agent-driven memory CLI maintains a persistent, source-grounded Wiki with citations and history. It is designed to let agents recall and evolve knowledge across sessions without embeddings or a vector database.
ccf / agentcairn
AgentCairn stores durable, cross-project memory as inspectable Markdown files with frontmatter, using an Obsidian vault as the source of truth. A local DuckDB cache supports hybrid search and retrieval, while MCP servers and host-specific plugins connect it to coding agents.
amanning3390 / flowstate-qmd
Flowstate-qmd anticipates what context an agent may need by watching session logs and prefetching it. It uses local SQLite-vec storage and GGUF models via node-llama-cpp, and provides an MCP-native server with automated configuration for major coding-agent clients.
yepyhun / Brainstack
Brainstack brings Hindsight, Graphiti, and MemPalace together under a single memory owner for Hermes. It prepares a purpose-built evidence packet for each query, using SQLite, Kuzu, and Chroma for state and retrieval management.
generalbusiness-ai / keep
Keep combines semantic search with BM25 full-text search and ranked graph traversal to retrieve stored information. It also indexes codebases, URLs, and multimedia with OCR support, while edge tags create navigable, bidirectional links between notes.
aliasocracy / hermes-tool-slimmer
Hermes Tool Slimmer reduces tool-schema token use through local keyword-based ranking and Tool Search support. It integrates as middleware without patching Hermes core on version 0.19 or later, and includes a dashboard for monitoring savings and managing tool configurations.
conorbronsdon / agent-context-os
Agent Context OS provides a Git-backed layer of plain Markdown context and workflows for Claude Code, Codex, and compatible coding agents. Its first-class adapters support multiple agents, while reviewed handoffs use hash-checked proposals and receipts.
Ladybug-Memory / hermes-memory-plugin
This local-only Hermes plugin stores memory in a single embedded .lbdb graph database file and supports BM25 keyword search and named graph relationships. Its importance-weighted recall brings higher-priority memories forward during context prefetch.
stepanov1975 / hermes-local-knowledge
Hermes Local Knowledge indexes configured local sources and selected Hermes runtime sources for capability discovery and routing. It registers five tools for search, retrieval, traversal, feedback, and usage reports, along with session and pre-LLM hooks.
MukundaKatta / hermes-agentmemory
Hermes AgentMemory is a pull-model episodic memory plugin with on-demand summarization and a trace.jsonl audit log of prefetch operations. It supports immediate event deletion, rather than retaining persistent summaries after deletion.
hermes-labs-ai / hermes-blind
Hermes Blind extracts the original goal from the first user turn of a long Claude Code or Codex session, with goals, first-sentence, and full extraction modes. It skips tool output and other non-user-turn records, and does not use a summarizing model or external service.
piprail / piprail
Piprail provides a TypeScript SDK for x402 payments and an MCP server for budget-bound agent wallets. It operates without a backend, database, or API keys.
yonro / hermes-xmemo-plugin
The XMemo plugin provides durable cross-session memory for Hermes, using lifecycle hooks for pre-turn recall and session-end snapshots. It uses a local SQLite layer for caching and write-outbox reliability, and can read memories from other connected agents.
futurebrowser / hermes-exabase-plugin
The Exabase plugin integrates Exabase M-1 for self-organizing long-term memory in Hermes. It provides tools for searching and manually storing memories, with configurable query expansion and result reranking.
sourcevault-ai / sourcevault-code-tools
SourceVault Code Tools provides private local code memory for repositories, with slash commands for search, file reads, context, and grounded Q&A. It supports Hermes CLI, Telegram, and Hermes Desktop, and includes repository sync and indexed Git-history search commands.
visorcraft / MongrelDB-Hermes
MongrelDB-Hermes provides long-term memory for Hermes Agent, with AES-256-GCM encryption at rest by default and native Rust FFI or HTTP daemon execution modes. Its retrieval options include HNSW approximate nearest-neighbor, sparse, and FM-index search.
penfieldlabs / hermes-penfield
Penfield provides memory management and knowledge graph traversal through 17 tools for Hermes Agent. It combines BM25, vector, and graph search, and can automatically checkpoint context during message compression.
conorbronsdon / agent-memory-kit
Agent Memory Kit is a file-based memory curation loop that works with Markdown and produces reviewable JSON diffs, without a database. Its “rot” and “lint” curators identify stale or conflicting facts, while a local Git repository tracks and can revert memory changes.