agentleFS
Sign inStart free
richkuorichkuo/rk-skills/skills/validate-fableplan-loop/SKILL.md

validate-fableplan-loop

Skill50 starsChanged 4 days agoMITView on GitHub

Use when the user asks to validate a GitHub issue (without Fable), conditionally plan it with fableplan, then autonomously drive it to a reviewed PR in one shot — "validate-fableplan-loop", "validate, plan, and work on #N", "validate and fableplan and fully automate #N". Runs validate-issue on your session model (not a Fable subagent), auto-applies its update-issue edits when the verdict calls for it, has fableplan produce and post a Fable 5.1 implementation plan (skipped only when the verdict's title-floored signal reads `plan: no`, which means the title score and the recomputed score are both below 75, with no safety flags and no `plan first: Yes` stamp in the issue's Execution block), then hands off to work-on-issue-loop — stopping instead when validation flags the issue as too large, architecturally infeasible, or already addressed by an existing PR, or recommended for closure as completed or invalid, or when the issue edit does not land. The non-Fable-validation counterpart to fable-validate-loop.

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
  1. Distinctive:

    Follow fable-validate-loop steps 1 through 6 with these changes:

    line 21

≈ “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

validate-fableplan-loop

Chain validate-issue → (conditional) update issue → (conditional) fableplan → work-on-issue-loop into one autonomous run.

This is fable-validate-loop with validation run through the plain validate-issue skill on your session model instead of a Fable 5.1 subagent.

Do not skip or reorder the chain. Validation gates planning, and the plan gates implementation. The only sanctioned skip is fable-validate-loop step 4's score gate, which step 4 below applies unchanged. Every other step of each skill still runs; only the "wait for the user's reply" moments are replaced by the decision rules in the cited steps.

Input

Same as fable-validate-loop, including the optional targetBranch.

Steps

Follow fable-validate-loop steps 1 through 6 with these changes:

Step 1 (validation): invoke the plain validate-issue skill (Skill tool, skill: validate-issue) instead of fable-validate. Let it run its full process — steps 0 through 8 — and produce the standard verdict block that fable-validate-loop step 1 quotes. Treat the verdict as structured output to parse yourself, and don't ask the user to confirm. Record the resolved issue as owner/repo#N per fable-validate-loop step 1; every later step passes that full reference and targets exactly this issue.

Step 2 (scope gate): fable-validate-loop step 2's STOP table applies, sourced from the plain validation (the already-addressing PR comes from validate-issue's step 1 linked-PR check; too large from validate-issue step 7). Rows (a) to (c) are Fable validation only and do not apply here.

Step 3 (update-issue edits): apply validate-issue's step 11 (this chain has no fable-validate step), from the current checkout (no worktree for issue edits, per validate-issue step 0), with fable-validate-loop step 3's read-back and STOP row unchanged. The stacked attribution line is Validated with LLM: <session model> | <session tier> | Harness: <harness> | validate-fableplan-loop, naming the session model and tier that ran the validation.

Step 4 (fableplan): fable-validate-loop step 4's score gate, safety carve-out, stamped plan flag, and top-band note apply, except that here the posted plan is the only Fable 5.1 stage: validation runs on the session model. When fableplan runs, give the planning subagent the validation findings (the verdict block and validate-issue's report) alongside the issue, and instruct fableplan to use the harness suffix validate-fableplan-loop in the posted comment's attribution footer.

Step 5 (handoff) applies unchanged — including that deviations follow work-on-issue step 2's plan-deviation policy and must each be named in the PR body.

Step 6 (report) applies unchanged. Cap the whole report at 55 words and 5 sentences, plain simple English in ASD-STE100 — apply the Response Style rules in CLAUDE.md/AGENTS.md.

Red Flags — STOP

fable-validate-loop's Red Flags table applies, reading "validate-issue" wherever it says "fable-validate", except the row for step 2 rows (a) to (c).

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.

Check your setup, free