clarify-spec
Skill1 starsChanged 34 days agoMITView on GitHub
Turn a software request and its source documents into established requirements, explicit assumptions, contradictions and decision questions before planning or when a specification gap appears. Does not choose missing product decisions, change tickets or implement code.
4 other projects use a version of this.
Adopted by
Most recently taken up by
- ZaxbyHub/opencode-swarmtheir own version, first seen Oct 30, 2026
- Greg-Klein/implementation-factoryword for word, first seen Oct 11, 2026
- zaxbysauce/opencode-swarmtheir own version, first seen Oct 8, 2026
- Svenja-dev/claude-code-skillstheir own version, first seen May 14, 2026
Other versions4 projects use a version of this.
Word for word · 1
- Greg-Klein/implementation-factory★ 1100% same
Edited · 0
None found.
Wrote their own version · 3
- zaxbysauce/opencode-swarm★ 519same name
- ZaxbyHub/opencode-swarm★ 518same name
- Svenja-dev/claude-code-skills★ 38same 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 repository13 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- collect-evidenceSkill
- document-changeSkill
- figma-reviewSkill
- gh-github-apiSkill
- gitlab-ticketsSkill
- glab-gitlab-apiSkill
- howSkill
- review-changeSkill
- self-checkSkill
- unslopSkill
- whySkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 16
Stop with useful partial findings when sources are missing.
≈ “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.
Files
Clarify a specification
Input: request, source documents, current explicit decisions and caller's source-precedence policy. Return established requirements with sources, out of scope, resolved contradictions and open decisions. Do not invent a universal document hierarchy if none is supplied.
- For each requested behavior, identify actors, inputs, states, outputs and failure conditions needed to implement it. Read relevant repository conventions and supplied references; mark inaccessible sources.
- Separate a contradiction from a detail omitted by a higher-level source. Apply the caller's precedence policy; preserve an explicit later decision and record the arbitration. Escalate a conflict whose intended authority remains unclear.
- Distinguish an established requirement, a sourced technical inference and an unresolved product choice. Existing conventions can settle an implementation detail, but cannot create permission rules, limits, endpoints or user-facing copy absent from the request.
- Ask only questions whose alternatives materially change the result, through the caller's interaction channel, and only once the sources, the repository and the linked work give no answer. An implementation, testing or measurement choice is a sourced technical inference to settle and record, never a question; a proposal the ticket makes while awaiting confirmation is implemented as proposed and flagged. Give concrete options and their consequences. Reuse answers already supplied. A question about what the system does today is a fact, not a decision: answer it yourself when it can be read or run without starting the application (an existing test, a script, a reachable endpoint). When only the running application can show it, ask, and mark the question
observablewith the measurement that will check the answer. - Propose testable acceptance wording without silently changing meaning. When a registry is supplied, preserve its ids and revision; propose additions as questions for its owner.
Stop with useful partial findings when sources are missing. Return questions to the caller rather than blocking unrelated work or answering on the user's behalf.
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.

