lock
Skill10 starsChanged 4 months agoMITView on GitHub
Convert vague input into a locked, executable target. Use when a task is fuzzy, over-broad, or stated as a wish. Returns the literal object, operation, output shape, real outcome, proof-of-done, smallest progress artifact, the highest-information missing variable, and the assumption to adopt if it stays missing.
2 other projects use a version of this.
Adopted by
Most recently taken up by
- devinat1/skillstheir own version, first seen Oct 13, 2026
- chris-santiago/ml-wikitheir own version, first seen Jul 16, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- devinat1/skills★ 1same name
- chris-santiago/ml-wiki★ 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 repository24 other files it gives its agents.
The file
Files
/ratchet:lock — scope lock
Ambiguity is where work goes to die. This command converts a wish into a target with edges. You do not begin work until the target is locked — but you also do not stall for missing information you can assume. Name the assumption and move.
Step 0 — Load state
ratchet statusProcedure
Interrogate the input until every field below has a concrete answer. Where the input is silent, infer the most reasonable value and mark it as an assumption.
- Literal object — the exact noun being operated on (this file, this prompt, this route, this dataset). Not a category. The specific thing.
- Operation — the single verb (build, fix, rank, migrate, decide, audit).
- Desired output shape — the concrete form of the result (a JSON file, a patch, a 3-row table, a 200-word spec). If you can't name the shape, the target isn't locked.
- Real outcome — what becomes true in the world when this is done. The reason, not the deliverable.
- Proof-of-done — the observable test that flips from fail to pass. One sentence, checkable by someone else.
- Smallest progress artifact — the tiniest thing you could produce in the next 20 minutes that is undeniably forward motion.
- Highest-information missing variable — the one unknown whose answer would change the plan most.
- Assumption if missing — the value you will adopt for that variable so work can start now, stated explicitly so it can be challenged later.
Output contract
OBJECT: <literal noun>
OPERATION: <single verb>
OUTPUT SHAPE: <concrete form>
REAL OUTCOME: <what becomes true>
PROOF-OF-DONE: <the checkable test>
SMALLEST ARTIFACT: <20-minute forward motion>
MISSING VARIABLE: <highest-information unknown>
WORKING ASSUMPTION: <adopted value + one-line justification>Serialize
ratchet state set objective "<operation> <object> → <output shape>, proof: <proof-of-done>"
ratchet state set phase lockMeter what follows
Once the target is locked, score its uncertainty so you spend only the loop it earns — don't reflexively run the full ratchet on a reversible one-liner:
ratchet score aperture '{"ambiguity":_,"terrain":_,"taste":_,"blastRadius":_,"reversibility":_}'- A0–A1 → go straight to
/ratchet:build, then/ratchet:verify. - A2–A3 → run
/ratchet:ignite's metered pipeline. - A4 → do not build yet; produce options and lock constraints first.
Then point forward: the locked target's smallest artifact is now the target of
/ratchet:build, or run /ratchet:auction first if the path to it is blocked.
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.

