agentleFS
Sign inStart free
neuraforge-cl-esneuraforge-cl-es/adaptive-engineering/skills/memory-grounding/SKILL.md

memory-grounding

Skill1 starsChanged 11 days agoMITView on GitHub

Ground durable project knowledge in exact repository evidence, detect stale claims when evidence changes, and reverify memory before it influences engineering decisions.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 6 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use repository evidence as the source of truth.

    line 7
  2. Distinctive:

    Read MEMORY.md before using persistent project memory.

    line 9
  3. Distinctive:

    Check that each claim's evidence still exists.

    line 14
  4. Very specific:

    Treat changed or contradicted claims as stale.

    line 16
  5. Distinctive:

    After implementation, refresh only affected claims.

    line 19
  6. Distinctive:

    Run the memory consistency gate before completion.

    line 20
  7. Distinctive:

    Never add a database, vector store, daemon, MCP server, or external service just to enable this skill.

    line 28

≈ “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.

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

Memory Grounding

Use repository evidence as the source of truth. Memory is only a compact, re-verifiable index over that evidence.

Read MEMORY.md before using persistent project memory.

Workflow

  1. Discover only claims relevant to the current task.
  2. Check that each claim's evidence still exists.
  3. Compare stored fingerprints with current evidence when possible.
  4. Treat changed or contradicted claims as stale.
  5. Inspect the source of truth before relying on stale or unverified knowledge.
  6. Persist only durable, reusable knowledge with precise evidence.
  7. After implementation, refresh only affected claims.
  8. Run the memory consistency gate before completion.

Mode behavior

  • FAST: verify directly relevant memory; avoid writing routine transient knowledge.
  • STANDARD: reverify planning claims and refresh affected durable knowledge.
  • DEEP: enumerate relevant claims and invariants, search for contradictions, and run a scoped drift check after implementation.

Never add a database, vector store, daemon, MCP server, or external service just to enable this skill. Plain repository-local files must remain sufficient.

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.

Check your setup, free