Refine Plan

Refine Plan

Loop the review pipeline over a planning artifact until no new findings are accepted. Writes back to the artifact file(s) in place. Supports plans, shells, and specs.

Task Tracking

At the start of every invocation (including re-runs from Step 5), use TaskCreate to create a task for each step:

• Resolve the artifact
• Run /review-plan skill
• Run /evaluate-findings skill
• Run /apply-findings skill
• Re-run /refine-plan skill if changed

Step 1: Resolve the Artifact

Determine Artifact Type

• Explicit argument — If the user specified a type (e.g., "refine plan", "refine shells", "refine spec"), use it
• Conversation context — Infer from conversation (e.g., if /draft-plan just ran, type is plan)
• Auto-detect — Check .turbo/ for existing artifacts. If multiple types exist, use AskUserQuestion

Resolve the Artifact File(s)

Plan

• Explicit path — use it
• Explicit slug — resolve to .turbo/plans/<slug>.md
• Single file — Glob .turbo/plans/*.md. If exactly one file exists, use it
• Most recent — most recently modified file
• Legacy fallback — .turbo/plan.md if .turbo/plans/ does not exist
• Nothing found — tell the user to run /turboplan (for a new task) or /pick-next-shell (for existing shells) and stop

Shells

• Explicit spec slug — Glob .turbo/shells/<slug>-*.md
• Explicit spec path — derive slug from filename, glob as above
• Single spec — Glob .turbo/specs/*.md. If exactly one, derive slug and glob for shells
• Most recent spec — most recently modified spec, derive slug and glob
• Nothing found — tell the user to run /draft-shells first and stop

For shells, read each shell file and extract from its YAML frontmatter: spec (source spec path) and depends_on. Verify the source spec exists. State the spec path, number of shells, and each shell's filename.

Spec

• Explicit path — use it
• Explicit slug — resolve to .turbo/specs/<slug>.md
• Single file — Glob .turbo/specs/*.md. If exactly one, use it
• Most recent — most recently modified
• Legacy fallback — .turbo/spec.md if .turbo/specs/ does not exist
• Nothing found — tell the user to run /draft-spec first and stop

If multiple candidates exist and the choice is non-obvious, use AskUserQuestion.

State the resolved path(s) before continuing.

Step 2: Run /review-plan Skill

Run the /review-plan skill on the resolved artifact.

Always run this step even if the artifact looks polished.

Step 3: Run /evaluate-findings Skill

Run the /evaluate-findings skill on the review findings from Step 2.

Step 4: Run /apply-findings Skill

Run the /apply-findings skill on the evaluated results.

Step 5: Re-run /refine-plan Skill if Changed

Check whether the artifact file(s) were edited during Step 4. Any edit counts.

The iteration number below refers to the /refine-plan run currently executing Step 5. It is not the iteration number of a prospective re-run. Iteration 1 is the initial run; iteration 2 is the first auto-re-run; iteration 3 is the second auto-re-run; iteration 4 and beyond exist only when the user opts in at the hard-cap ask. Iterations 1 and 2 always follow the classification gate (they never trigger the hard cap at their own Step 5, even when the auto-re-run they spawn would be iteration 3). The hard cap fires at the end of iteration 3 and every iteration thereafter.

Iterations 1 and 2, if changes were made, classify what Step 4 edited:

• Structural edits — run /refine-plan again via the Skill tool, passing the artifact type and resolved path. If the round contains both structural and prose-only edits, treat it as structural and re-run automatically.
• Prose-only edits only (reworded sentences in place, fixed stale examples, clarified existing text without changing meaning) — output a summary of what changed, then use AskUserQuestion to ask whether to run one more round or stop here. Do not silently continue or silently stop.

Iterations 1 and 2, if changes were made but you believe re-running is unnecessary, use AskUserQuestion to ask for skip permission. Do not skip silently.

Judge whether another round is worthwhile by the trend across iterations: when rounds have stopped surfacing defects (contradictions, infeasible steps, missing requirements) and keep surfacing improvements of kinds earlier rounds already applied, the loop has converged even though the edits were structural — recommend stopping. A round that surfaces no defects is the termination signal; never add a confirmation round, an extra reviewer, or review steps beyond this skill's own.

Judge the kind of surviving defect as well as the trend. Once earlier rounds have drained the design- and requirement-level defects and a round's findings are dominated by claims that a named identifier does not exist or does not match its declaration, the artifact has passed the point where reviewing it as text pays, even when the count is rising and every finding is genuine. Implementation surfaces that class immediately, so recommend stopping. Expect any round that adds mechanism to seed defects at the seams it creates.

When a round's findings turn on how a platform, framework, or dependency behaves and nobody has observed that behavior, prefer a cheap instrumented experiment to another review round. Run it in a temp directory outside the repo, and record what it shows in the artifact before judging whether another round is warranted.

Iteration 3 or later, if Step 4 of this run made changes, the hard cap is reached. This replaces the classification gate above for iteration 3 and every iteration after it. Output a summary of what is still changing and whether it is structural or prose-only. Then use AskUserQuestion to offer three options: continue for another iteration, stop here and accept the artifact as-is, or escalate to /consult-oracle for a different perspective on the remaining issues.

When the same class of defect recurs across iterations, stop patching the individual instance and instead write the root-cause invariant into the artifact itself, enumerating the worked failures it must prevent. In the same pass, re-read the whole artifact against the new invariant and fix every instance it catches, including text written before the invariant existed. Treat recurrence on a new axis of the same invariant as a signal that the invariant is incomplete: widen it to cover the new axis rather than assuming the latest fix failed.

The re-invocation is a full, fresh run of this skill. Every step (1-5) executes with its own task tracking and skill invocations.

Then use the TaskList tool and proceed to any remaining task.

Structural Edit Examples by Type

• Plan — added or removed steps, new or removed design decisions, rewired dependencies between steps, changed testing strategy
• Shells — added or removed shells, changed Produces/Consumes/Covers Spec Requirements wiring, changed frontmatter depends_on, added or removed spec requirement coverage
• Spec — added or removed sections, new or removed requirements, rewired cross-references, changed acceptance criteria

Rules

• Every step must run in every iteration. /evaluate-findings is a judgment gate that must run before /apply-findings touches the artifact. Each step must invoke its designated skill via the Skill tool.
• Re-invocations from Step 5 are full runs with fresh task tracking and complete skill invocations.
• The artifact file(s) are the only files that should change. For shells, do not modify the source spec.