done
Post-Task Verification (/done)
This skill is MANDATORY after every task. No exceptions. Not optional. Not "nice to have". REQUIRED.
Even for single-line changes, one-file edits, or "trivial" fixes — you MUST run /done before marking any task as complete.
Workflow
Step 1: Type Check
Run /fix-ts-errors — this runs pnpm type-check across all packages and loops until zero errors.
If the task only touched a specific app, you may scope the first pass (--filter=frontend or --filter=backend), but always run the full check at least once.
Do not proceed to Step 2 until type-check passes clean.
Step 2: Parallel Code Review
Run /parallel-review — this launches code-reviewer + coderabbit review agents in parallel on the changed code.
If issues are found:
- Fix critical and serious issues
- Re-run
/fix-ts-errorsafter fixes - Report what was fixed
Step 3: Simplify
Run /simplify — this reviews changed code for reuse, quality, and efficiency.
If improvements are suggested:
- Apply the improvements
- Re-run
/fix-ts-errorsafter changes
Step 4: Verify Correctness
- Review the final code logic to verify it does what was intended
- If tests exist for the changed code, run them
- Ask yourself: "Would a staff engineer approve this?"
Step 5: Report
Briefly report:
- What was checked
- Any issues found and fixed during review
- Final status: clean or remaining concerns
Step 6: Suggest Commit Messages
After everything passes, suggest two ready-to-paste git commit commands. Do NOT run them — only print them for the user to choose.
Use conventional commit format (feat:, fix:, refactor:, chore:, docs:).
Option A — Detailed:
git commit -m "feat: add user role validation to auth middleware
- Add role-based guard to /api/admin routes
- Update UserRole type with new 'manager' variant
- Add validation tests for all role combinations"
Option B — One-liner:
git commit -m "feat: add user role validation to auth middleware"
Rules for commit message suggestions:
- Analyze the actual
git diffto understand what changed - The subject line should explain WHY, not WHAT (the diff shows what)
- Detailed version uses
-bullet points for key changes (3-5 bullets max) - Use the correct prefix:
feat:for new features,fix:for bugs,refactor:for restructuring,chore:for tooling/config,docs:for documentation - Keep subject line under 72 characters
- Never commit automatically — only suggest
Rules
- NEVER skip this skill — it runs after every task, period
- NEVER mark a task as complete without running /done first
- NEVER say "the changes are small so I'll skip review" — small changes cause big bugs
- If a step finds issues, fix them before proceeding to the next step
- If you already ran
/fix-ts-errorsduring implementation, run it again — new issues may have been introduced
More from bhagyamudgal/skills
review-pr
Deep, anti-slop review of a GitHub PR. Grounds findings in the linked issue's intent, runs a Claude reviewer (+ conditional silent-failure hunter) in parallel with existing CodeRabbit comments fetched from the PR, then critic-passes the findings before printing. Use when user says "review this pr", pastes a GitHub PR URL, or asks "check this pull request". NOT for local uncommitted changes — use /parallel-review for those.
12fix-pr-review
Triage and fix CodeRabbit / review-pr findings on a GitHub PR, then reply + resolve conversations. Classifies each finding (FIX/DISMISS/DEFER/DISAGREE/NEEDS-INPUT), runs /done, posts specific replies. Use on CodeRabbit review URLs, PR URLs, or local review files. Pairs with /review-pr.
7forge-plan
Use when starting any non-trivial feature or design task that needs the full idea-to-implementation pipeline. Also use when user says "forge", "forge plan", "full pipeline", or wants the complete AI development workflow before shipping.
6harden-plan
Pre-code quality gate that runs /review-pr's anti-slop lens against a written plan BEFORE any code is written. Grounds the plan against the real codebase, runs 11 category checks (security, concurrency, round-trip, control-flow, error-handling, pattern-consistency, plus /review-pr's Q1-Q6), then grills the user one question at a time until the plan is hardened. Use when the user says "harden my plan", "check my plan", "grill my plan before I code", "lint this plan", or invokes `/harden-plan` explicitly. Also invoke proactively after `/superpowers:brainstorm` or `/grill-me` completes with a written plan and before any implementation begins. Do NOT invoke after coding has started — redirect to `/review-pr` / `/fix-pr-review` in that case.
6project-discovery
Deep project discovery and architecture planning. Use when starting a new project, when user says "new project", "let's build", or asks for help architecting/planning a non-trivial application.
5qa
Use when user says "qa", "test the flow", "QA this", "verify the UI", "test this page", "check if it works", "end-to-end check", or after completing code changes that affect UI and need manual verification replaced by automation.
4