.agents/skills/joycraft-implement/SKILL.md
Execute atomic specs with TDD — read spec, write failing tests, implement until green, hand off to session-end
npx skillsauth add maksutovic/joycraft joycraft-implementInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
You have one or more atomic spec paths to execute. Your job is to implement each spec using strict TDD — tests first, confirm they fail, then implement until green.
The user should provide one or more spec paths (e.g., docs/specs/my-feature/add-widget.md).
If no spec path was provided, tell the user:
No spec path provided. Check
docs/specs/for available specs, or provide a path like:$joycraft-implement docs/specs/feature-name/spec-name.md
For each spec path:
Specs are designed to be self-contained, but if you need more context:
> **Parent Brief:** line). Read it for broader feature context.docs/specs/<feature-name>/ where the feature name is derived from the brief filename (strip the date prefix and .md — e.g., 2026-04-06-token-discipline.md → token-discipline).This is not optional. Write tests FIRST.
Using the spec's Test Plan:
Walk through every Acceptance Criterion in the spec:
If any criterion is not met, keep implementing. Do not move on until all criteria are green.
Check the spec's Edge Cases table. For each scenario:
If the user provided multiple specs:
Spec [name] complete. [N] specs remaining.
Tip: Run /new before starting the next spec. Your artifacts are saved to files — this conversation context is disposable.
When all specs are implemented and passing:
Implementation complete:
- Spec(s): [list spec names] — all Acceptance Criteria met
- Tests: [N] written, all passing
- Build: passing
Next steps:
- Run $joycraft-session-end to capture discoveries and wrap up
Tip: Run /new before starting the next step. Your artifacts are saved to files — this conversation context is disposable.
development
Spawn an independent verifier subagent to check an implementation against its spec -- read-only, no code edits, structured pass/fail verdict
development
Assess and upgrade your project's AI development harness — score 7 dimensions, apply fixes, show path to Level 5
testing
Wrap up a session — capture discoveries, verify, prepare for PR or next session
development
Produce objective codebase research by isolating question generation from fact-gathering — subagent sees only questions, never the brief