
git
by tao3k
WIP-designed to bridge the gap between human intent and machine execution. Built on a robust Tri-MCP architecture, it physically separates cognitive planning (The Brain), atomic execution (The Hands), and surgical coding (The Pen) to ensure context hygiene and safety.
SKILL.md
name: "git" version: "2.0.0" description: "Git integration with LangGraph workflow support, Smart Commit V2, and Spec-Awareness" routing_keywords: [ # Core verbs (high priority) "git", "commit", "push", "pull", "merge", "rebase", "checkout", "stash", "tag", # High-frequency phrases "commit code", "save changes", "commit changes", "push code", "save work", "check in", "submit code", "version control", "branch", "repo", "repository", "history", "diff", "status", "log", "hotfix", "pr", "pull request", "code review", ] intents: [ "hotfix", "pr", "branch", "commit", "stash", "merge", "revert", "tag", "status", ] authors: ["omni-dev-fusion"] permissions:
- "filesystem:*"
- "terminal:run_command"
- "knowledge:ingest"
Git Skill
Code is Mechanism, Prompt is Policy
Architecture
This skill uses @skill_command decorator in scripts/*.py files.
Commands are automatically exposed via MCP as git.command_name.
Available Commands
| Command | Description |
|---|---|
git.status | Show working tree status |
git.stage_all | Stage all changes (with security scan) |
git.commit | Commit staged changes |
git.smart_commit | Smart Commit workflow (stage → scan → approve → commit) |
git.push | Push to remote |
git.log | Show commit logs |
Smart Commit Workflow
Use git.smart_commit for secure, human-in-the-loop commits:
# Step 1: Start workflow
git.smart_commit(action="start")
# Returns workflow_id and diff preview
# Step 2: After LLM analysis and user approval
git.smart_commit(action="approve", workflow_id="xxx", message="feat: description")
Flow: stage_and_scan → route_prepare → format_review → re_stage → interrupt → commit
Staged Files Feature
Stage and Scan Workflow
The stage_and_scan function provides automatic staging with security validation:
Stage All Files → Security Scan → Lefthook Pre-commit → Finalize
Key Features
-
Automatic Staging
stage_and_scan(project_root=".") # Returns: {staged_files, diff, security_issues, lefthook_error} -
Security Scanning
- Detects sensitive files (
.env*,*.pem,*.key,*.secret, etc.) - Automatically un-stages detected files
- Returns list of security issues
- Detects sensitive files (
-
Lefthook Integration
- Runs pre-commit hooks after staging
- Re-stages files modified by lefthook formatters
- Returns lefthook output for review
Staged Files Commands
| Command | Description |
|---|---|
git.stage_all() | Stage all changes with security scan |
git.status() | Show staged files and working tree status |
git.diff() | Show staged diff |
Security Patterns Detected
.env*, *.env*, *.pem, *.key, *.secret, *.credentials*
id_rsa*, id_ed25519*, *.priv
secrets.yml, secrets.yaml, credentials.yml
Usage Guidelines
Read Operations (Safe - Use Claude-native bash)
git status
git diff --cached
git diff
git log --oneline
Write Operations (Use MCP Tools)
| Operation | Tool |
|---|---|
| Stage all | git.stage_all() (scans for secrets) |
| Commit | git.commit(message="...") |
| Push | git.push() |
| Smart Commit | git.smart_commit(action="start") |
Key Principle
Read = Claude-native bash. Write = MCP tools.
Score
Total Score
Based on repository quality metrics
SKILL.mdファイルが含まれている
ライセンスが設定されている
100文字以上の説明がある
GitHub Stars 100以上
1ヶ月以内に更新
10回以上フォークされている
オープンIssueが50未満
プログラミング言語が設定されている
1つ以上のタグが設定されている
Reviews
Reviews coming soon


