agentleFS
OneDro1dOneDro1d/dark-factory/skills/scope-retire/SKILL.md

scope-retire

Skill4 starsChanged 2 months agoApache-2.0View on GitHub

Archive a completed or abandoned agent-notepad while keeping its journal history intact. Marks the notepad retired (SCOPE.md status + a final journal milestone), stops it from auto-loading and syncing on future sessions, and preserves the append-only sessions/ history and handoffs for later recall — it never deletes the journal, the memory index entries, or the remote. Reversible: a later scope-init re-activates it. Use when an objective is done, a scope is being wound down, or a notepad should be shelved. Triggers on "scope-retire", "retire this scope", "archive the notepad", "close out this objective", "wind down the scope".

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 3 are distinctive, and 3 are 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.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  • could be automatic: a setting could enforce it
  1. Distinctive:

    Read SCOPE.md done-criteria and report which are met vs open, so the operator retires with eyes open.

    line 26
  2. Very specific:

    Do one last rewrite of NOTES.md (§6.1): set Current goal to a one-line outcome ("Retired <date>: <done | abandoned because …>"), Next …

    line 31
  3. Very specific:

    Add a status line at the top of SCOPE.md, e.g.

    line 50
  4. Distinctive:

    Do not remove the hooks or the settings file — retirement is reversible.

    line 61
  5. Very specific:

    If a remote exists, do one best-effort git -C <notepad> add -A && commit && push of the retirement snapshot (final NOTES.md, journal …

    line 65
  6. Distinctive:

    No destructive ops: no journal deletion, no memory-index purge, no repo/remote deletion, no force-push.

    line 83

≈ “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

scope-retire — archive a notepad, keep the history

Retiring a notepad means "stop treating this as an active objective" — not "delete it". The journal (sessions/*.jsonl), the handoffs, the memory-index (wing=<prefix>) entries, and the git remote all stay. Retirement is a reversible status change plus a clean final snapshot. See DESIGN.md (§6 file contracts, §16 acceptance).

Never destructive. This skill does not rm journals, does not empty the memory index, does not delete the GitHub repo, and does not force-push over history. If the operator wants hard deletion, that is a separate manual action they perform themselves — surface it, don't do it.


Procedure

1. Confirm the notepad and its state

  • Resolve the notepad directory (cwd or the operator-named scope). Confirm it is a notepad (has SCOPE.md + repos.manifest.json + sessions/).
  • Read SCOPE.md done-criteria and report which are met vs open, so the operator retires with eyes open. Retiring with open criteria is allowed (abandonment) — just say so.

2. Write a final snapshot into NOTES.md

Do one last rewrite of NOTES.md (§6.1): set Current goal to a one-line outcome ("Retired <date>: <done | abandoned because …>"), Next action to "none — scope retired", and leave the Key refs / Blockers as the honest final state. This is the summary a future reader lands on first; keep it terse and un-redacted of caveats.

3. Append a retirement milestone to the journal (do NOT rewrite history)

Append one new event to the current session journal (sessions/<ISO8601>_<id>.jsonl) — never edit or delete existing lines:

{"ts":"<now>","kind":"milestone","text":"scope retired: <outcome>","refs":[],"commit":null,"session":"<id>"}

The append-only journal is the history of record; retirement adds a closing entry, it does not erase what came before.

4. Mark SCOPE.md retired

Add a status line at the top of SCOPE.md, e.g. > **Status:** RETIRED <date> — <one-line outcome>. This is the durable, human-visible flag that the notepad is shelved. Leave Objective + Done-criteria in place for the record.

5. Disarm auto-load / auto-sync for future sessions

So a stray future session in this directory doesn't resurrect it as "active":

  • In the notepad's .claude/settings.json, the read/sync hooks key off notepad detection; a RETIRED SCOPE.md is the signal the SessionStart hook should treat as "inject the final NOTES.md snapshot read-only, skip the digest rebuild and best-effort push". Record the retired status where the hook can see it (SCOPE.md status line is the source of truth).
  • Do not remove the hooks or the settings file — retirement is reversible.

6. Final sync, then leave the remote alone

If a remote exists, do one best-effort git -C <notepad> add -A && commit && push of the retirement snapshot (final NOTES.md, journal milestone, SCOPE status) so the archived state is durable off-machine. Then stop — do not delete the remote branch or repo. Preserve handoffs/ as-is.


Done when

SCOPE.md shows RETIRED with an outcome, NOTES.md holds the final snapshot, the journal has its closing milestone appended (nothing removed), future sessions won't treat it as active, and — if there's a remote — the archived state is pushed. Report: the notepad path, the outcome (done vs abandoned + which done-criteria were open), and that history was preserved (journal + handoffs + memory index untouched).

Guardrails

  • Journal is append-only: add the milestone, never edit or delete prior lines.
  • No destructive ops: no journal deletion, no memory-index purge, no repo/remote deletion, no force-push. Hard deletion is the operator's own manual call.
  • Reversible: a later scope-init on the same directory re-activates the notepad.

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.