agentleFS
wuzishu3-webwuzishu3-web/agent-memory-vault/SKILL.md

agent-memory-vault

Skill0 starsChanged 5 months agoMITView on GitHub

Persistent shared long-term memory + knowledge ingest for AI coding agents over a Markdown/Obsidian vault. Use when you need cross-session memory, want to remember decisions/projects/preferences, ingest an external source (article/video/repo/transcript) into durable notes with automatic cross-references, or search what was learned before. Triggers on "remember this", "save to memory", "what did we decide about X", "ingest this source", "boot memory", or starting work that should persist across sessions.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. youcisla/Agent-Foundry★ 1same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repositoryOne other file it gives its agents.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read what it prints (hot cache + recent log + task hits) before acting.

    line 25
  2. Very specific:

    Write a durable note (see _system/WRITE_GUIDE.md for type → location):

    line 27
  3. Distinctive:

    Never write secrets (API keys, tokens, passwords, cookies) or private/raw chat logs.

    line 58
  4. Very specific:

    Don't hand-write source pages — route external sources through ingest.py.

    line 60

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

agent-memory-vault

A shared long-term memory system over a Markdown vault at $AGENT_MEMORY_VAULT (default ~/agent-memory-vault). All scripts live in this skill's scripts/.

When to use

  • Starting meaningful work → boot, so you're oriented by prior context.
  • A durable fact emerged (decision / project progress / user preference / verified conclusion) → write it.
  • You processed an external source (read an article, analyzed a video/repo, fetched a page) worth keeping → ingest it.
  • You need prior context ("did we solve this before?", "what did we decide about X?") → query.
  • Ending a session with state changes → update the hot cache.

Core commands

Boot at the start of substantive work:

python3 scripts/agent_memory_boot.py --agent <agent-name> --task "<one line>"

Read what it prints (hot cache + recent log + task hits) before acting. Current user instructions always override stored memory.

Write a durable note (see _system/WRITE_GUIDE.md for type → location):

python3 scripts/agent_memory.py --type <decision|knowledge|project|playbook|...> \
  --agent <agent-name> --title "<topic phrase, no date>" --summary "<one line>" --body "<...>"

Ingest an external source — do NOT hand-write a source note; use this:

python3 scripts/ingest.py --agent <agent-name> \
  --title "<topic>" --summary "<YOUR high-quality summary>" \
  --body-file <notes.md> --url "<source url>" --source-type <article|video|repo|transcript>

It writes the page, propagates bidirectional [[cross-references]], runs a quality gate (sub-par input is quarantined to the inbox, never the source folder), and rebuilds indexes. The summary must be your own in-context analysis — the script only does bookkeeping. Never let it auto-summarize.

Search:

python3 scripts/query_vault.py "<question>"

End-of-session, if state changed:

python3 scripts/update_hot.py --agent <agent-name> --last "<what you did>" \
  --active "<remaining>" --pending "<blocked on>" --dont "<pitfall to avoid>"

Rules

  1. Long-term value only — ephemeral chatter does not belong in the vault.
  2. Never write secrets (API keys, tokens, passwords, cookies) or private/raw chat logs.
  3. Uncertain facts get status: 待核验 and a note of what needs verifying.
  4. Don't hand-write source pages — route external sources through ingest.py.
  5. WRITE_GUIDE.md is the single authority for what content goes where.

Maintenance (optional, schedulable)

python3 scripts/vault_health_check.py     # lint: stale/overdue/orphan pages
python3 scripts/dream_cycle.py            # consolidation pass
python3 scripts/build_embeddings.py       # rebuild vector index (incremental)
python3 scripts/extract_relations.py      # rebuild relation graph

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.