Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Git commit patterns, formats, and conventions for GSD methodology. Provides atomic commits per task, structured commit messages, planning file commits, branch management, and milestone tag operations.
.claude/skills/a5c-ai-git-integration/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 262% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 2% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 67% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 51% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 54% | 0% |
Each task in a plan produces one atomic commit:
bash# After completing task 1 of PLAN-1.md git add src/auth/oauth.ts src/auth/types.ts git commit -m "feat(auth): implement OAuth2 login endpoint Phase 72, Plan 1, Task 1: Implement login endpoint - Added POST /api/auth/login with email/password validation - Returns JWT token on success, 401 on invalid credentials - Rate limited to 5 attempts per minute Refs: Phase-72/PLAN-1/Task-1"
Commit message format:
<type>(<scope>): <short description>
Phase <N>, Plan <M>, Task <K>: <task title>
- <change 1>
- <change 2>
Refs: Phase-<N>/PLAN-<M>/Task-<K>Types: feat, fix, refactor, test, docs, chore, style, perf.
Planning files are committed separately from code:
bash# After generating plans git add .planning/phase-72/PLAN-1.md .planning/phase-72/PLAN-2.md git commit -m "plan(phase-72): create execution plans Phase 72: Authentication - PLAN-1.md: OAuth2 login and token management (4 tasks, wave 1) - PLAN-2.md: Auth middleware and role guards (3 tasks, wave 1) Refs: Phase-72/planning"
Planning commit types: plan, state, research, context, verify, audit.
Generate commit messages from task context:
Input:
phase: 72
plan: 1
task: 1
title: "Implement login endpoint"
type: "feat"
scope: "auth"
changes: ["POST /api/auth/login endpoint", "JWT token generation", "Rate limiting"]
Output:
"feat(auth): implement OAuth2 login endpoint
Phase 72, Plan 1, Task 1: Implement login endpoint
- Added POST /api/auth/login endpoint
- JWT token generation on success
- Rate limiting (5 attempts/minute)
Refs: Phase-72/PLAN-1/Task-1"Create annotated tags for milestone completions:
bashgit tag -a v1.0 -m "Milestone v1.0: MVP Launch Phases completed: 70, 71, 72, 73 Requirements covered: R1-R11 Audit result: PASSED Key deliverables: - Project scaffolding and CI pipeline - Database layer with migrations - OAuth2 authentication - REST API endpoints"
Detect current branch state:
json{ "branch": "main", "clean": false, "staged": ["src/auth/oauth.ts"], "unstaged": [".planning/STATE.md"], "untracked": ["src/auth/oauth.test.ts"], "ahead": 3, "behind": 0 }
Separate .planning/ files from source code files:
bash# Stage code changes only git add src/auth/oauth.ts src/auth/types.ts # Stage planning changes only (separate commit) git add .planning/phase-72/SUMMARY.md .planning/STATE.md
Rules:
Analyze commit history for progress tracking:
Phase 72 commits:
abc1234 feat(auth): implement OAuth2 login endpoint (Task 1)
def5678 feat(auth): add token refresh mechanism (Task 2)
ghi9012 feat(auth): implement logout and revocation (Task 3)
jkl3456 test(auth): add auth endpoint tests (Task 4)
Total: 4 commits, 12 files changed, +450 -30 linesGenerate diff summaries for SUMMARY.md:
Changes summary:
Files created: 4
Files modified: 2
Files deleted: 0
Lines added: 450
Lines removed: 30
New files:
src/auth/oauth.ts (120 lines)
src/auth/tokens.ts (85 lines)
src/auth/types.ts (45 lines)
src/middleware/auth.ts (90 lines)
Modified files:
src/routes/index.ts (+15 -2)
package.json (+3 -0)Bash to run git status to see current changesBash to stage only the files related to the current task: git add <files>Bash to commit: git commit -m "<message>"git log -1Bash to identify planning file changes: git status .planning/git add .planning/<files>Read to load milestone audit data for tag annotationBash to create annotated tag: git tag -a <version> -m "<annotation>"git tag -l <version>Bash to run git status --porcelainBash to run git rev-list --count HEAD..@{upstream} for behind countBash to run git rev-list --count @{upstream}..HEAD for ahead countexecute-phase.js - Atomic commit per task during executionverify-work.js - Commit verification results and fixesquick.js - Atomic commit for quick task executiondebug.js - Commit debug fixescomplete-milestone.js - Create milestone git tagadd-tests.js - Commit each test file atomicallyresearch-phase.js - Commit research documentplan-phase.js - Commit generated plans (planning commit)json{ "operation": "commit|tag|status|history|diff", "status": "success|error", "commit": { "hash": "abc1234", "message": "feat(auth): implement OAuth2 login endpoint", "filesChanged": 3, "linesAdded": 120, "linesRemoved": 5 }, "tag": { "name": "v1.0", "annotation": "Milestone v1.0: MVP Launch" }, "branchStatus": { "branch": "main", "clean": true, "ahead": 0, "behind": 0 } }
| Setting | Default | Description | |---------|---------|-------------| | autoCommit | true | Automatically commit after each task | | commitMessageFormat | conventional | Commit message format (conventional/simple) | | separatePlanningCommits | true | Separate planning from code commits | | signCommits | false | GPG-sign commits | | tagPrefix | v | Prefix for milestone tags |
| Error | Cause | Resolution | |-------|-------|------------| | Nothing to commit | No changes to stage | Verify task made expected changes | | Merge conflict | Conflicting changes in working tree | Resolve conflicts before committing | | Dirty working tree | Uncommitted changes from previous task | Commit or stash before proceeding | | Tag already exists | Milestone tag already created | Use --force or choose different tag name | | Detached HEAD | Not on a branch | Checkout a branch before committing |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,381 | 5,241 | -3% | 1 | 1 | 0% | 1,101 | 2,584 | +135% | 0 | 0 | — |
case-02 | fail→fail | 4,472 | 5,719 | +28% | 1 | 1 | 0% | 938 | 2,521 | +169% | 0 | 0 | — |
case-03 | fail→fail | 9,182 | 5,729 | -38% | 1 | 1 | 0% | 1,528 | 2,994 | +96% | 0 | 0 | — |
case-04 | fail→pass | 4,665 | 2,724 | -42% | 1 | 1 | 0% | 733 | 2,657 | +262% | 0 | 0 | — |
case-05 | fail→fail | 6,144 | 3,446 | -44% | 1 | 1 | 0% | 1,177 | 2,830 | +140% | 0 | 0 | — |
case-06 | fail→fail | 9,197 | 3,150 | -66% | 1 | 1 | 0% | 1,692 | 2,622 | +55% | 0 | 0 | — |
case-07 | fail→pass | 13,562 | 2,353 | -83% | 1 | 1 | 0% | 2,406 | 2,446 | +2% | 0 | 0 | — |
case-08 | fail→pass | 8,369 | 1,192 | -86% | 1 | 1 | 0% | 1,353 | 2,253 | +67% | 0 | 0 | — |
case-09 | fail→fail | 6,851 | 3,618 | -47% | 1 | 1 | 0% | 1,258 | 2,763 | +120% | 0 | 0 | — |
case-10 | fail→fail | 2,918 | 2,369 | -19% | 1 | 1 | 0% | 398 | 2,386 | +499% | 0 | 0 | — |
case-11 | fail→pass | 10,832 | 4,347 | -60% | 1 | 1 | 0% | 1,960 | 2,956 | +51% | 0 | 0 | — |
case-12 | fail→pass | 9,051 | 3,887 | -57% | 1 | 1 | 0% | 1,853 | 2,861 | +54% | 0 | 0 | — |
case-13 | fail→fail | 3,941 | 2,105 | -47% | 1 | 1 | 0% | 659 | 2,407 | +265% | 0 | 0 | — |
case-14 | fail→fail | 4,558 | 1,241 | -73% | 1 | 1 | 0% | 794 | 2,256 | +184% | 0 | 0 | — |
case-15 | fail→pass | 7,620 | 999 | -87% | 1 | 1 | 0% | 1,299 | 2,207 | +70% | 0 | 0 | — |
case-16 | fail→pass | 7,799 | 1,540 | -80% | 1 | 1 | 0% | 1,310 | 2,255 | +72% | 0 | 0 | — |
case-17 | pass→pass | 6,329 | 1,747 | -72% | 1 | 1 | 0% | 1,035 | 2,373 | +129% | 0 | 0 | — |
case-18 | fail→pass | 7,717 | 5,535 | -28% | 1 | 1 | 0% | 1,395 | 3,036 | +118% | 0 | 0 | — |
case-19 | fail→fail | 8,935 | 6,483 | -27% | 1 | 1 | 0% | 1,911 | 3,291 | +72% | 0 | 0 | — |
case-20 | fail→fail | 8,569 | 2,110 | -75% | 1 | 1 | 0% | 1,320 | 2,430 | +84% | 0 | 0 | — |
case-21 | fail→fail | 9,517 | 7,703 | -19% | 1 | 1 | 0% | 2,018 | 3,581 | +77% | 0 | 0 | — |
case-22 | fail→fail | 8,439 | 6,834 | -19% | 1 | 1 | 0% | 1,594 | 3,192 | +100% | 0 | 0 | — |
case-23 | fail→fail | 8,222 | 8,568 | +4% | 1 | 1 | 0% | 1,511 | 3,382 | +124% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted, and 21 counted toward the lift figure. The other 2 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +35 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.