agentleFS
omdenkar2-uiomdenkar2-ui/all-in-one/skills/aio-autoplan/SKILL.md

aio-autoplan

Skill0 starsChanged 33 days agoMITView on GitHub

Automatically refine and independently review an executable backend plan, with scope and developer-interface checks before engineering review. Use for autoplan or automatic backend planning; return to an authorized build controller without starting another pipeline.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are 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.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use AIO Spec, Architecture and Plan for missing contracts, critical failure traces, exact interfaces, owned units and verification.

    line 16
  2. Distinctive:

    Use the stage contract. Return readiness to the caller.

    line 25

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

AIO Autoplan

Produce a ready backend plan from the accepted task. This adapts gstack's ordered planning orchestration: settle scope, check affected developer interfaces, then review engineering last. Standalone Autoplan delivers planning artifacts; an authorized Complete Destruction run continues implementation after the gates pass.

Inherit the resource contract, original journal and source authority. Standalone runs resolve thorough once. Select only relevant stages; Autoplan is a planning coordinator inside the current run, not another end-to-end controller. Do not launch separate reviewers for each planning lens or mandatory outside-model voices.

Protocol

  1. Bind the exact accepted task, repository and current changes. Reuse an existing matching contract/plan. Identify materially missing facts and continue independent planning while obtaining them. Accepted implementation authority persists through planning; a planning request alone does not authorize a build.
  2. Resolve the scope mode. Default HOLD SCOPE. Use Brainstorm only for unresolved outcomes or alternatives. SCOPE EXPANSION, SELECTIVE EXPANSION and SCOPE REDUCTION control proposals; additions or deferrals outside the accepted task need a scope decision before entering R/U/V records. Settled requirements remain binding.
  3. Use AIO Spec, Architecture and Plan for missing contracts, critical failure traces, exact interfaces, owned units and verification. A bounded prototype is conditional on a consequential uncertainty; it consumes the same journal and cannot be mistaken for production proof.
  4. For an affected public API, SDK, CLI or operator integration path, apply Devex Review to the proposed developer journey before final engineering review. Default DX POLISH. Planned examples and inferred friction are planning evidence; unavailable runtime behavior stays untested. Findings become explicit plan amendments, not a new implementation run.
  5. Freeze the final candidate and run the structural plan gate. Then reserve one independent PLAN reviewer through Plan Review, covering scope/requirement consistency, relevant developer-interface findings, feasibility and adversarial failure handling. Engineering review is last: a subsequent material amendment invalidates affected readiness and needs a bounded scoped re-review. Reuse a matching current-source review instead of repeating it.
  6. Resolve confirmed plan blockers in the shared corrective-wave allowance. Never reset caps or create independent per-lens retries. If required review, facts or remaining capacity are unavailable, retain a resumable blocked plan with the exact next action. Structural validity alone cannot mark it ready.

Return and handoff

Return the human plan, machine index, R/U/V coverage, selected scope/DX modes, actual amendments, gate results and separate independent feasibility/adversarial verdicts. Unchosen opportunities remain a short optional list outside accepted scope. Avoid duplicate full reports for each lens.

Use the stage contract. Return readiness to the caller. Under an authorized build, the existing controller continues with AIO Work; a standalone planning invocation ends with its reviewable plan.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.