Hermes Atlas
Memory & knowledge · works with Hermes Agent

claude-mem

thedotmack/claude-mem

Persistent memory that records agent sessions, compresses them with AI and feeds context into later sessions

In short

claude-mem is a persistent memory system that records what a coding agent does during a session, compresses it with AI, and injects relevant context into later sessions. Its repository description lists Hermes among the agents it works with, next to Claude Code, OpenClaw, Codex, Gemini, Copilot and OpenCode.

What claude-mem does

claude-mem captures tool usage observations while an agent works, generates semantic summaries, and makes them available to future sessions so the agent keeps continuity on a project after a session ends or reconnects. It was built for Claude Code first and installs through npx claude-mem install, with an --ide flag for targets such as OpenCode, Antigravity, Pi and T3 Code. A separate installer script covers OpenClaw gateways.

Memory is searchable through a mem-search skill, and a web viewer shows the memory stream in real time at the worker URL printed on startup. Privacy tags exclude sensitive content from storage, and context configuration controls what gets injected. After sign-in, the installer lets you pick a memory provider: the claude-mem observer, your own OpenRouter or Gemini key, or your Anthropic plan. The opening part of the README shows install flows for several agents but no Hermes-specific command.

Key features

  • Captures tool usage observations during agent sessions and writes semantic summaries
  • Injects relevant context from earlier sessions into new ones
  • Progressive disclosure with token cost visibility for layered retrieval
  • A mem-search skill for querying project history
  • Web viewer UI that streams memory in real time
  • Privacy tags that exclude sensitive content from storage

When to use it

  • Keeping a coding agent aware of earlier decisions across separate sessions
  • Searching what an agent did on a project last week
  • Running one memory layer across several coding agents

Who it is for: Developers who use coding agents such as Claude Code or Codex and want sessions to carry context forward.

How it fits with Hermes Agent

The repository description names Hermes as a supported agent, but claude-mem is a general memory tool aimed first at Claude Code, and the opening part of the README documents install steps for other agents rather than Hermes.

How to install claude-mem

These commands are copied from the project's README. Check the repository for the latest steps before you run them.

npx claude-mem install

Note: The claude-mem observer starts with a 30-day free trial and then falls back to your Anthropic plan unless you subscribe, and an explicit --provider flag skips the sign-in.

FAQ

What is claude-mem?

claude-mem is a persistent memory system for coding agents. It records tool usage during sessions, summarizes it with AI, and injects relevant context into future sessions.

Does claude-mem work with Hermes Agent?

The repository description says it works with Hermes among other agents, including Claude Code, OpenClaw, Codex, Gemini, Copilot and OpenCode. The opening part of the README shows install flows for several of those but no Hermes-specific command, so check docs.claude-mem.ai before relying on it.

How do I install claude-mem?

Run npx claude-mem install. The README warns that npm install -g claude-mem installs only the SDK and library, without registering plugin hooks or the worker service.

Similar memory for Hermes Agent

All memory

Related guides: SOUL.md for Hermes Agent: what it is and how to write one · Run multiple Hermes agents with profiles