Skip to content
Get packs

Add Mnemon Persistent Memory

Gives NanoClaw Claude agents graph-based memory that persists across container restarts

Source Download skill
Repository
nanocoai/nanoclaw
Stars
31,072
Licence
MIT
Files
4 files

What it does

This skill adds a memory layer to agents running in NanoClaw containers. It installs the mnemon tool into the agent container image and registers Claude Code hooks so the agent recalls relevant past context before answering and stores new insights after each turn. Memory is saved in the agent group's .claude folder so it survives container restarts.

Great for

  • Developers running NanoClaw agent groups on the default Claude provider who want conversations to carry over between sessions
  • Teams who want to inspect or reset what an agent remembers by deleting a folder on the host

Skip it if

  • Agent groups using a non-Claude provider such as opencode, because mnemon hooks only fire under the Claude Code target and will never run for those groups
  • Users handling sensitive information who have not reviewed what gets stored, since memory persists on disk and is surfaced back into future prompts

Try asking

  • “Add persistent memory to my NanoClaw agents so they remember past conversations”
  • “Install mnemon in the agent container and wire it into the entrypoint”
  • “Check whether mnemon hooks are registered in my running agent container”
  • “Reset the memory for one of my agent groups”
Step-by-step guide: how to use Add Mnemon Persistent MemoryInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 4 original files, the MIT licence and an attribution note.

    add-mnemon.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip add-mnemon.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip add-mnemon.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Run the provider check grep first so you do not apply the skill to groups that will never use the hooks
  • Pin MNEMON_VERSION to a release you have reviewed rather than trusting the latest tag automatically
  • Review the stored memory folder periodically, since it persists and can contain sensitive context

What's inside

  • mnemon-dockerfile.test.ts
  • mnemon-entrypoint.test.ts
  • REMOVE.md
  • SKILL.md

Redistributed unmodified from nanocoai/nanoclaw under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.