richkuo fable-validate-fableplan-loop
Chain fable-validate → (conditional) update issue → fableplan → work-on-issue-loop into one autonomous run: Fable 5.1 validates the issue, the main agent fixes the issue description if needed, Fable 5.1 plans the implementation (plan posted to the issue), and work-on-issue-loop implements the plan and drives the PR through review to convergence.
This is fable-validate-loop with the score gate removed — the only difference is that fableplan always runs, regardless of the validated complexity score; there is no "skip when the score is below 71" rule. Reach for this when even simple issues should get a posted, Fable-vetted plan before implementation (e.g. the plan comment doubles as documentation, or the repo's simple-looking issues have a history of hiding traps). If skipping the plan for lower-band issues is fine, use fable-validate-loop instead — it's the cheaper default.
Do not skip or reorder the chain. Validation gates planning (a plan built on refuted claims is wrong), and the plan gates implementation (that's the point of routing through fableplan). This variant has no sanctioned skip: every step runs, and only the "wait for the user's reply" moments are replaced by the decision rules in the cited steps.
Input
Same defaults as fable-validate: issue URL, #<N> / <N> / owner/repo#N, or nothing (defaults to the latest open issue in the current repo).
Optional targetBranch (orchestration form { issue, targetBranch } or a prose "target branch "): passed unchanged to every validate, plan, and build step in the chain, so the baseline and the PR base are that branch instead of the repo default. work-on-issue step 1 ("Target") owns its validation.
Steps
Follow fable-validate-loop steps 1 through 6 with these changes:
Step 1 (fable-validate) applies unchanged; it produces the standard verdict block:
**#<N>: Update issue description? <Yes|No>** · Complexity: <score>/100 — Capability <k> (Risk <r>, Uncertainty <u> — <driver>); Volume <v> (Scope <s>, Coupling <c>, Verification <x>) · fableplan: <yes|no> · Scope: <OK | too large — split/umbrella/narrow>
Step 2 (scope gate) applies unchanged — the same four STOP conditions:
| Condition | Action |
|---|---|
Scope: too large (split / umbrella / narrow flagged) |
STOP. Report the disposition and proposed parts — splitting is a human call. |
| Architecture marked ❌ Infeasible | STOP. Report the infeasibility and the "Optimal direction" note. |
| A merged PR already implements the fix | STOP. Report the PR and the close/repurpose recommendation. |
| An open PR is already addressing the issue | STOP. Report the overlapping PR; supersede/join/wait is a human call. |
Step 3 (update-issue edits): apply them per fable-validate step 5 / validate-issue step 11; the stacked Validated with LLM: … attribution line uses the harness suffix fable-validate-fableplan-loop.
Step 4 (fableplan): there is no score gate — fable-validate-loop's score gate, safety carve-out, and top-band note do not apply; fableplan runs for every issue that passed the step-2 scope gate, whatever the validated score. Instruct fableplan to use the harness suffix fable-validate-fableplan-loop in the posted comment's attribution footer, so the comment records the actual entry point. The rest of the step applies unchanged: fableplan's planning-phase-only invocation, and the validation verdict handed to the planner (verified/refuted claims, the Optimal-direction note when architecture was ⚠️, 5c concerns), with the scratchpad kept for step 5.
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, 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. Read its "only sanctioned skip is the step-4 score gate" wording as "no sanctioned skip at all": a score below 71 never skips fableplan here.