autopilot-refine
Skill9 starsChanged 4 days agoMITView on GitHub
Use when documents need factual, structural, stylistic, or review-driven correction; preserve history. Owns revision; default preset. Not for new documents, empirical work, or evaluation/result reporting (lab).
In this repository27 other files it gives its agents.
- AGENTS.mdAGENTS.md
- analyze-projectSkill
- analyze-userSkill
- artifact-projection-readSkill
- auditSkill
- autopilot-applySkill
- autopilot-codeSkill
- autopilot-designSkill
- autopilot-draftSkill
- autopilot-labSkill
- autopilot-researchSkill
- autopilot-shipSkill
- autopilot-specSkill
- code-executeSkill
- code-planSkill
- code-refineSkill
- code-reportSkill
- code-testSkill
- design-componentsSkill
- design-handoffSkill
- design-initSkill
- design-refsSkill
- design-reviewSkill
- design-tokensSkill
- draft-refineSkill
- draft-strategySkill
- session-tidySkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 3 are distinctive, and 1 is 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.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 24
Use only this router, harness-manifest.json, and core/WORKFLOW.md §0.2 to propose a route.
- Distinctive: line 37
After approval, read the portable procedure set:
- Distinctive: line 41
Read the complete portable procedure files listed here when they apply.
- Distinctive: line 50
Follow the portable artifact, worktree, role, and verification guards in the selected owner contract.
≈ “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
autopilot-refine
This is a compact pre-approval entry router. Its manifest-owned frontmatter is the authoritative discovery metadata; it intentionally contains no execution procedure.
Pre-approval boundary
Use only this router, harness-manifest.json, and core/WORKFLOW.md §0.2 to
propose a route. Present the one-time confirmation in core/WORKFLOW.md §0.4
unless the same route and scope are already approved. Do not load references
before approval.
Post-approval owner contract
Main session only: after approval, read capabilities/autopilot-refine.md. Registered owners read only capabilities/autopilot-refine/references/owner-execution.md; stage workers read only their assigned contracts.
The other entry-router steps apply to the main session. A registered owner uses
only its Route steps and the latest receipt for runtime commands.
Reference Index
After approval, read the portable procedure set:
Shared Procedure References
Read the complete portable procedure files listed here when they apply; this logical read-set is shared across Claude, Codex, and OpenCode.
capabilities/autopilot-refine/references/examples-and-constraints.mdcapabilities/autopilot-refine/references/owner-execution.mdcapabilities/autopilot-refine/references/process-stages.mdcapabilities/autopilot-refine/references/versioning-and-modes.md
Guard pointer
Follow the portable artifact, worktree, role, and verification guards in the selected owner contract. Before the first durable capability artifact, compile and bind the checked route even for direct intensity; direct is a compiled inline node, not route absence. A restriction on native subagents/agents is surface-local and never selects direct execution. Runtime projections must report unsupported mechanics and must not claim physical instruction masking, token, billing, or cost savings without verified evidence.
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.