agentleFS
claude-worldclaude-world/claude-agent/.claude/skills/memory-manager/SKILL.md

memory-manager

Skill5 starsChanged 8 months agoMITView on GitHub

Consolidate, prune, and organize memory files autonomously. Auto-triggered every 30 minutes via /loop. Also use when user says "clean up memory", "what do you remember", or "memory status".

11 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 2 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 8 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read all files in memory/:

    line 27
  2. Distinctive:

    Merge duplicates: If the same fact appears in multiple places, keep only the most complete version

    line 36
  3. Very specific:

    If learned-today.md has entries from a previous day, move them to memory/archive/YYYY-MM-DD-learned.md

    line 41
  4. Very specific:

    If active-threads.md has threads with no activity > 48h, mark them as resolved

    line 42
  5. Distinctive:

    If archive files are older than 30 days, overwrite them with empty content using Write (no Bash/Delete tool available)

    line 43

“Common” is measured: kept by at least 6 in 10 of the 8 edited versions of this file we could read. “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

Memory Manager

You are the memory management skill for Claude-Agent. Your job is to keep the memory system healthy and organized.

Steps

1. Audit Current Memory

Read all files in memory/:

  • MEMORY.md — index file
  • user-profile.md — user identity
  • active-threads.md — conversation threads
  • pending-tasks.md — tasks and reminders
  • learned-today.md — daily learnings (if exists)
  • contacts.md — known contacts (if exists)

2. Consolidate

  • Merge duplicates: If the same fact appears in multiple places, keep only the most complete version
  • Update timestamps: Mark stale entries (threads with no activity > 24h)
  • Resolve completed tasks: Move [x] tasks to a "Recently Completed" section, then archive after 3 days

3. Archive

  • If learned-today.md has entries from a previous day, move them to memory/archive/YYYY-MM-DD-learned.md
  • If active-threads.md has threads with no activity > 48h, mark them as resolved
  • If archive files are older than 30 days, overwrite them with empty content using Write (no Bash/Delete tool available)

4. Update Index

Update MEMORY.md with:

  • Current file count
  • Last consolidation timestamp
  • Brief summary of active state (X threads, Y tasks, Z contacts)

5. Report

Output a brief status:

Memory Status:
- User profile: [complete/partial/empty]
- Active threads: N
- Pending tasks: N
- Contacts: N
- Archive entries: N
- Last consolidated: [timestamp]

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.