agentsclimarketplace

Verify

Skill Sagargupta16/craftsmanship/skills/verify

Skills that encode engineering discipline -- plan, guard, verify, review, audit, ship. Compatible with 45+ AI agents via skills.sh.

Install
npx -y skills add Sagargupta16/craftsmanship --skill verify

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when completing a task, before claiming work is done, or before creating a PR. Enforces verification beyond "tests pass" -- actually run the code, hit the endpoints, open the browser. Prevents premature completion claims.

SKILL.md

4.1 KB, 941 tokens by cl100k_base, as published. Nobody here has run it

Verify Before Claiming Done

The Rule

Never claim work is complete based solely on:

  • Code looking correct
  • Type checks passing
  • Linter being clean
  • Tests passing (tests verify code correctness, not feature correctness)

Always verify the actual feature works as intended.

Quick Reference

Change TypeMinimum Verification
Backend APIRun server, hit endpoints, check responses
Frontend UIStart dev server, test in browser, check edge cases
Database migrationRun migration, verify schema, test rollback
Infrastructure/IaCPlan/validate, check for destructive changes
Bug fixReproduce bug, apply fix, confirm fix, check for regressions
RefactorRun full test suite, verify behavior unchanged
DocumentationBuild docs, check links, verify accuracy

Backend Verification

1. Run the server / service
2. Hit affected endpoints (curl, httpie, or test client)
3. Check response status codes and bodies
4. Verify error cases return proper errors
5. Check logs for warnings or errors
6. Run the test suite
7. Check for N+1 queries or performance issues

Example:

# Start server
uvicorn main:app --reload

# Test happy path
curl -X POST http://localhost:8000/api/users \
  -H "Content-Type: application/json" \
  -d '{"name": "test", "email": "[email protected]"}'

# Test error case -- should return 422
curl -X POST http://localhost:8000/api/users \
  -H "Content-Type: application/json" \
  -d '{"name": ""}'

Frontend Verification

1. Start dev server
2. Navigate to affected pages
3. Test the golden path (happy path)
4. Test edge cases:
   - Empty state
   - Error state
   - Loading state
   - Long text / overflow
   - Mobile viewport
5. Check browser console for errors
6. Test keyboard navigation (accessibility)
7. Run test suite and build

Database Migration Verification

1. Run migration forward
2. Verify schema matches expectations
3. Run migration backward (rollback)
4. Verify rollback is clean
5. Run migration forward again
6. Seed data if needed, verify queries work
7. Run application test suite

Infrastructure Verification

1. Run terraform plan / cdk diff
2. Review changes -- especially destroys
3. Check for state drift
4. Verify no secrets in plan output
5. Apply to staging/dev first
6. Verify resources created correctly
7. Run smoke tests against new infrastructure

Bug Fix Verification

1. Reproduce the original bug (confirm it exists)
2. Apply the fix
3. Verify the bug is fixed (same reproduction steps)
4. Check for regressions (related features still work)
5. Write a test that would have caught this bug
6. Run full test suite

Pre-PR Checklist

CheckCommand
Tests passpytest / pnpm test / go test ./...
Build succeedspnpm build / cargo build
No type errorspyright / tsc --noEmit
No lint warningsruff check / eslint .
No committed secretsSearch for API keys, tokens, passwords
Branch is up to dategit fetch origin && git log origin/main..HEAD
Changes are scopedgit diff --stat -- only expected files changed

Anti-Patterns

Anti-PatternProblemDo Instead
"Tests pass so it works"Tests verify code, not featuresTest the actual feature manually
"It compiles so it's done"Compilation != correctnessRun and verify behavior
"I checked the happy path"Bugs hide in edge casesTest error states, empty states, boundaries
Skipping verification for small changesSmall changes cause big bugsAlways verify, even one-liners
Only verifying in isolationIntegration issues missedTest in context of the full application
Claiming done without running itMost common failure modeRun it, use it, then claim done

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.