/craft

Craft

Task

$ARGUMENTS

Orchestration

You are the orchestrator for the full Praxis craft pipeline. Invoke each skill in turn, pass each skill's output as input to the next, and check in with the user between stages.

Pipeline

clarifying-intent → [slicing-stories] → sketching-design → driving-tdd
  → code-reviewing → code-improving → verifying-and-adapting

1. clarifying-intent — Pass the task. The skill returns one of:

   • A trivial change statement → implement it directly and stop.
   • A Story-Level Behavioral Spec → confirm with the user, then go to step 3.
   • A Feature Brief → confirm with the user, then go to step 2.
   • Open questions → answer them with the user, then re-invoke.

2. slicing-stories — Pass the Feature Brief. The skill returns a slice map. Confirm with the user, pick the first slice, and re-invoke clarifying-intent with that slice's story to produce a Story-Level Behavioral Spec for it.

3. sketching-design — Pass the spec. The skill returns a design sketch, marks itself skipped, or surfaces a spec issue. If a spec issue is surfaced, return to clarifying-intent.

4. driving-tdd — Pass the spec and (if produced) the sketch. The skill commits implementation as it goes, then returns the AC checklist, feedback log, and session summary. If it surfaces ## Feedback, return to clarifying-intent for that gap, then re-invoke driving-tdd.

5. code-reviewing — Pass the spec and the TDD session summary. The skill returns a review report.

6. code-improving — Pass the review report and the spec. The skill commits fixes and returns an improvement summary. If it surfaces ## Feedback, return to clarifying-intent.

7. verifying-and-adapting — Pass the spec, AC checklist, feedback log, session summary, optional sketch, and (multi-slice only) slice map. The skill returns a verification summary, optionally an updated spec, and a routing recommendation. Act on the recommendation:

   • Next slice → pick the next slice from the slice map and return to step 3 with a fresh clarifying-intent for it.
   • Rework → return to step 4 with the gaps.
   • Escalate → return to step 1 at feature level.
   • Done → the story (or feature) is complete.

Notes

• Stop at every user checkpoint to confirm scope and any surfaced findings.
• Skills accept either inline content or a path/handle to read. Pass whatever is convenient.
• If you choose to persist any artifact (brief, slice map, spec, sketch, review, summary) to disk, you decide where. There is no enforced layout.