memory-extract
Skill1 starsChanged 8 months agoMITView on GitHub
Extract durable facts, decisions, preferences, and conventions from the current conversation and save them to appropriate memory files. Use after complex discussions, debugging sessions, architecture decisions, or when valuable information was discovered that should persist across sessions. Triggers on "save this to memory", "remember this", "extract memories", "don't forget", "note this for later".
In this repository6 other files it gives its agents.
- code-navigateSkill
- code-searchSkill
- code-simplifySkill
- magic-docsSkill
- memory-organizerSkill
- session-summarySkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 41
Prefer updating existing entries over creating duplicates
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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.
Files
Memory Extract — Save Valuable Information to Persistent Memory
Analyze recent conversation and extract information worth persisting.
What to Extract
| Value | Examples | Destination |
| High | Architecture decisions, working rules, safety rules, project conventions | MEMORY.md |
| High | New ports, services, paths, domains | MEMORY.md or TOOLS.md (confirm first) |
| High | User preferences, communication style | USER.md (confirm first) |
| Medium | Bug root causes, deployment steps, config discoveries | memory/YYYY-MM-DD.md |
| Skip | Trivial Q&A, dead-end debugging, generic knowledge, duplicates | — |
Process
- Scan the conversation for decisions, new facts, rules, and solved problems.
- Check MEMORY.md and
memory/listing for duplicates. Skip anything already recorded. - Write:
- MEMORY.md / TOOLS.md / USER.md → show the diff, get confirmation before writing
- memory/YYYY-MM-DD.md → write directly (low risk)
- Append, never overwrite. Use bullet points.
- Report what was extracted and where:
Extracted N items: - MEMORY.md: 2 entries (rules, environment) - memory/2026-03-31.md: 3 entries (debug findings) - Skipped 1 duplicate
Rules
- Concise bullet points, not prose
- Include dates when rules were established
- Max 10 items per extraction — prioritize by value
- Prefer updating existing entries over creating duplicates
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
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.