run-skill-generator
Skill2 starsChanged 55 days agoApache-2.0View on GitHub
Discover, verify, and persist a project-specific recipe that lets future agents build, launch, and drive the real application surface from a clean environment.
In this repository6 other files it gives its agents.
- AGENTS.mdAGENTS.md
- chatcut-production-mirror/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- runSkill
- verifySkill
- html-artifactsSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
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
- Very specific: line 8
Create a durable run-<unit> skill for the smallest meaningful application unit.
- Distinctive: line 12
Search the repo root and relevant app/package directories.
- Distinctive: line 16
Use package/build manifests, developer docs, existing scripts, environment examples, container config, and executable entrypoints.
≈ “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
Generate a project run skill
Create a durable run-<unit> skill for the smallest meaningful application unit. Document commands that were actually exercised.
Existing recipe first
Search the repo root and relevant app/package directories. If an existing run skill targets the same unit, preserve what works and repair stale or missing steps.
Discover the recipe
Use package/build manifests, developer docs, existing scripts, environment examples, container config, and executable entrypoints. Determine prerequisites, build/install, launch/readiness, how to drive the real surface, evidence capture, and cleanup.
Prove it
Execute the recipe from a fresh enough process with an explicit working directory and isolated runtime state. Drive at least one meaningful surface action. Repair the recipe until the documented commands match what actually works.
Persist
Default location:
<unit>/.claude/skills/run-<unit-name>/
SKILL.md
driver.*Include the working directory, verified prerequisites, build/install, launch, readiness, drive commands/tooling, evidence method, cleanup, and observed gotchas. Keep local credentials and machine-specific absolute home paths out of the skill.
Finally, follow the generated skill literally in a new process where practical. Only report success after paths resolve, launch reaches readiness, the real surface can be driven, and cleanup works.
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.