agentleFS
pantheon-orgpantheon-org/tekhne/skills/agentic-harness/vault-consolidate/SKILL.md

vault-consolidate

Skill10 starsChanged 39 days agoMITView on GitHub

Propose or apply consolidation of episodic memories into semantic ones, combining related episodic captures into higher-level semantic summary notes. Use after a long session or when the user asks to clean up, organise, summarise, merge, or declutter memories, or requests memory cleanup or memory management. Always run --propose first so the user can review before --apply commits changes to the vault.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    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:

    After a long session with many captures

    line 26
  2. Distinctive:

    When the user asks to "consolidate", "summarise", or "clean up" memories

    line 27
  3. Distinctive:

    Always propose before applying:

    line 32
  4. Very specific:

    If --propose outputs "No episodic memories found to consolidate", skip Step 3 — there is nothing to apply.

    line 45

≈ “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

vault-consolidate

Cluster related episodic memories and synthesise them into durable semantic notes.

Mindset

Think of --propose as generating a diff and --apply as committing the merge. You never commit without reviewing the diff first. The human is always the final gatekeeper — the agent's job is to surface candidates, not to decide what gets promoted to long-term memory.

When to use

  • After a long session with many captures
  • When the user asks to "consolidate", "summarise", or "clean up" memories
  • Periodic maintenance (e.g. end of week)

Workflow

Always propose before applying:

# Step 1 — generate proposals (written to vault inbox for review)
vault-cli consolidate --propose [--project <id>]

# Step 2 — open vault inbox, review, set status: approved or rejected
# (edit 00-inbox/consolidation-proposals.md in Obsidian)

# Step 3 — apply approved proposals
vault-cli consolidate --apply

If --propose outputs "No episodic memories found to consolidate", skip Step 3 — there is nothing to apply.

What happens

  • --propose: clusters episodic memories by semantic similarity, calls the inference command to synthesise each cluster, writes proposals to 00-inbox/consolidation-proposals.md
  • --apply: reads the proposals file, writes approved ones as semantic notes, marks source episodics as superseded, logs rejected ones to audit

Important

Source episodic memories are never deleted — only marked superseded. Human edits in Obsidian are always respected and never overwritten.

Never

  • Never run --apply without first running --propose — doing so bypasses human review and irreversibly promotes unreviewed clusters to semantic memory
  • Never run --apply without confirming the user has reviewed the proposals — always pause at Step 2 and wait for the user to confirm before proceeding
  • Never delete episodic source memories manually — they are preserved automatically and serve as audit history

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.