菜单
Automated git workflow helpers for common development tasks like creating feature branches, cleaning up merged branches, and interactive rebasing. Use when the user mentions git branching, branch cleanup, feature workflow, or git automation. No prerequisites required - uses native git commands.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/jakenuts/agent-skills --skill git-workflow命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
基于 SOC 职业分类
| name | git-workflow |
| description | Automated git workflow helpers for common development tasks like creating feature branches, cleaning up merged branches, and interactive rebasing. Use when the user mentions git branching, branch cleanup, feature workflow, or git automation. No prerequisites required - uses native git commands. |
| compatibility | Requires git installed (available in most development environments) |
| metadata | {"requires-setup":false} |
Automated helpers for common git workflows and branch management tasks.
This skill uses native git commands and bash scripts - no additional tools or SDKs needed. Works immediately in any environment with git installed.
Create a new feature branch following naming conventions:
# Interactive: prompts for feature name
cd /path/to/repo
bash ~/.claude/skills/git-workflow/scripts/create-feature-branch.sh
# Non-interactive: specify feature name
bash ~/.claude/skills/git-workflow/scripts/create-feature-branch.sh "add-user-auth"
# Creates: feature/add-user-auth
Remove local branches that have been merged to main/master:
cd /path/to/repo
bash ~/.claude/skills/git-workflow/scripts/cleanup-merged-branches.sh --dry-run
bash ~/.claude/skills/git-workflow/scripts/cleanup-merged-branches.sh
bash ~/.claude/skills/git-workflow/scripts/cleanup-merged-branches.sh --force
Simplified interactive rebase with common options:
cd /path/to/repo
# Rebase last 3 commits
bash ~/.claude/skills/git-workflow/scripts/interactive-rebase.sh 3
# Rebase since a specific commit
bash ~/.claude/skills/git-workflow/scripts/interactive-rebase.sh abc123
# Rebase onto main
bash ~/.claude/skills/git-workflow/scripts/interactive-rebase.sh main
All scripts are located in the skill's scripts/ directory and can be run with bash.
| Script | Purpose | Usage |
|---|---|---|
create-feature-branch.sh | Create feature branches with naming conventions | bash script.sh [feature-name] |
cleanup-merged-branches.sh | Clean up merged branches | bash script.sh [--dry-run|--force] |
interactive-rebase.sh | Interactive rebase helper | bash script.sh <commit-count|ref> |
# 1. Ensure main is up to date
git checkout main
git pull origin main
# 2. Create feature branch
bash ~/.claude/skills/git-workflow/scripts/create-feature-branch.sh "user-authentication"
# 3. Work on feature...
# (make commits)
# 4. Push when ready
git push -u origin feature/user-authentication
# 1. Update main branch
git checkout main
git pull origin main
# 2. Preview what will be deleted
bash ~/.claude/skills/git-workflow/scripts/cleanup-merged-branches.sh --dry-run
# 3. Clean up merged branches
bash ~/.claude/skills/git-workflow/scripts/cleanup-merged-branches.sh
# Squash last 5 commits into one
bash ~/.claude/skills/git-workflow/scripts/interactive-rebase.sh 5
# In the editor that opens:
# - Change 'pick' to 'squash' (or 's') for commits 2-5
# - Keep first commit as 'pick'
# - Save and close
# Then edit the combined commit message
All scripts include:
No additional tools, SDKs, or environment variables required.
Create and iterate on shareable visual diagrams (architecture maps, flowcharts, state machines, class/domain models, mindmaps, dependency maps) via the Agent Vision MCP server or its HTTP action API. Use when the user asks for a diagram, visualization, or "draw this out" during a conversation, or when the user hands you an Agent Vision app URL + connection token.
ElevenLabs Speech Engine — the SDK for giving a voice interface to a custom agent you host yourself (your existing chat agent, LangGraph workflow, containerized LLM service, etc.). ElevenLabs runs the audio pipeline (STT, turn detection, VAD, TTS, barge-in); your server receives transcripts over a WebSocket and streams text replies back. Use when wiring an existing non-ElevenLabs agent to voice, when the user mentions "Speech Engine", `seng_*` engine IDs, `@elevenlabs/elevenlabs-js` `speechEngine.attach()` / `SpeechEngine.Server`, Python `AsyncElevenLabs.speech_engine`, or full-duplex voice with bring-your-own-LLM. Also covers the matching client SDKs (`@elevenlabs/client`, `@elevenlabs/react`, `@elevenlabs/react-native`, the convai widget) for the browser/mobile side. Do NOT use this skill for ElevenAgents (fully hosted agent platform — different product, see distinction below) or plain TTS / STT calls.
WolverineFx (.NET) framework for in-process command/event handling, asynchronous messaging, durable inbox/outbox, sagas, and HTTP endpoints. Use when working in a .NET project that references `WolverineFx*` packages, calls `UseWolverine()`, or whenever the user asks about Wolverine handlers, message routing, transactional outbox, transports (RabbitMQ, Azure Service Bus, SQS, Kafka, NATS, Pulsar, PostgreSQL, SQL Server, Redis, MQTT, etc.), Wolverine.HTTP endpoints, Marten/EF Core/RavenDb integration, sagas/process managers, middleware, cascading messages, side effects, FluentValidation integration, tracked-session testing, or migrating to/from MediatR/NServiceBus/MassTransit. Also applies to Critter Stack and JasperFx conversations that mention Wolverine.