janitor
Skill0 starsChanged 13 days agoMITView on GitHub
Read-only vault scan that points at what needs doing — literal templates, unfiled inbox, old-format notes, path-broken links, a rogue duplicate folder, a vanished skill, a broken feature contract — each routed to the skill that fixes it. Use for "what needs tidying?" or as the engine behind the morning check and weekly review.
4 other projects use a version of this.
Adopted by
Most recently taken up by
- udaysharmadev/Skillstheir own version, first seen Sep 17, 2026
- gonun13/agent-skillstheir own version, first seen Sep 5, 2026
- Nathan5674312/fatetheir own version, first seen Sep 4, 2026
- adamjali/janitor-templatetheir own version, first seen Jun 28, 2026
Other versions4 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 4
- udaysharmadev/Skills★ 12same name
- adamjali/janitor-template★ 1same name
- gonun13/agent-skills★ 0same name
- Nathan5674312/fate★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository14 other files it gives its agents.
- captureSkill
- capture-to-inboxSkill
- escalate-to-architectSkill
- file-noteSkill
- interview-meSkill
- migrateSkill
- process-inboxSkill
- recallSkill
- resumeSkill
- scanSkill
- selftestSkill
- taskSkill
- weekly-reviewSkill
- learn-acceleratorSkill
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
1
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 22
Run: python $MYOS_ENGINES/myos_janitor_scan.py (full punch-list) or --brief (one line per issue).
- Distinctive: line 34
Never mass-fix. Route what the user OKs.
- Very specific: line 40
After INTENTIONALLY adding/removing a skill, run myos_janitor_scan.py --bless so the expected-skills manifest matches (else it flags the …
“Common” is measured: kept by at least 6 in 10 of the 3 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.
Files
janitor
The vault's manager, not its worker. Scans read-only for drift and emits a punch-list, each item tagged with the skill that fixes it. It never edits notes — it points; the named skills do the work. A morning health brief consumes --brief; the weekly review reads the full list. Background: [[NOTES]].
When to Use
- "What needs tidying / what's drifting in the vault?"
- As the detector behind a morning health brief (terse) and
weekly-review(reflective).
Procedure
- Run:
python $MYOS_ENGINES/myos_janitor_scan.py(full punch-list) or--brief(one line per issue). Add--escalateon a scheduled run to auto-queue CRITICAL feature breakages. - Relay the ⚠ / 🚨 lines — each already names its owning skill (🚨 = critical, feature-breaking).
- Hand each issue to the named skill (do not do the work here):
- un-rendered
<% tp… %>→migrate(it renders the supported subset) - unfiled inbox →
process-inbox/file-note - no-
type:notes or a leftover legacy header →migrate - path-broken
[[links]]→escalate-to-architect - rogue duplicate folder (a second "Inbox") →
escalate-to-architect - a missing/expected skill → restore from git, or
escalate-to-architect - invalid SKILL.md YAML → fix the frontmatter (quote any value containing
:) - 🚨 broken feature contract (a template edit severed a coupling listed in
contracts.json) →escalate-to-architect;--escalatealready appended it toEscalations.md. Add a contract whenever a new template↔code coupling is introduced. - memory index too big / memory duplicated →
escalate-to-architect/myos_memory_sync.sh
- un-rendered
- Never mass-fix. Route what the user OKs; leave the rest on the list.
Pitfalls
- Read-only. Never edits notes. The janitor points; skills act — this prevents a runaway auto-fix across the vault. (Its only writes: seeding the skills manifest, and appending to
Escalations.mdwith--escalate.) - Bare
[[Concept]]links are NOT flagged — graph-first concept nodes are intentional; only path-style breaks surface. - The no-
type:list is a backlog, not an emergency — route a few per pass, don't dump 150 edits at once. - After INTENTIONALLY adding/removing a skill, run
myos_janitor_scan.py --blessso the expected-skills manifest matches (else it flags the change).
Verification
- Output lists each issue with a count + owning skill; the scan makes zero writes to notes;
--blessseeds/updates the expected-skills manifest.
Cadence
Daily (--brief --escalate, scheduled) + weekly (weekly-review, full). On demand anytime.
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.