Agent skill
commit-work
Create high-quality git commits: review/stage intended changes, split into logical commits, and write clear commit messages (including Conventional Commits). Use when the user asks to commit, craft a commit message, stage changes, or split work into multiple commits.
Install this agent skill to your Project
npx add-skill https://github.com/davila7/claude-code-templates/tree/main/cli-tool/components/skills/productivity/commit-work
SKILL.md
Commit work
Goal
Make commits that are easy to review and safe to ship:
- only intended changes are included
- commits are logically scoped (split when needed)
- commit messages describe what changed and why
Inputs to ask for (if missing)
- Single commit or multiple commits? (If unsure: default to multiple small commits when there are unrelated changes.)
- Commit style: Conventional Commits are required.
- Any rules: max subject length, required scopes.
Workflow (checklist)
- Inspect the working tree before staging
git statusgit diff(unstaged)- If many changes:
git diff --stat
- Decide commit boundaries (split if needed)
- Split by: feature vs refactor, backend vs frontend, formatting vs logic, tests vs prod code, dependency bumps vs behavior changes.
- If changes are mixed in one file, plan to use patch staging.
- Stage only what belongs in the next commit
- Prefer patch staging for mixed changes:
git add -p - To unstage a hunk/file:
git restore --staged -porgit restore --staged <path>
- Prefer patch staging for mixed changes:
- Review what will actually be committed
git diff --cached- Sanity checks:
- no secrets or tokens
- no accidental debug logging
- no unrelated formatting churn
- Describe the staged change in 1-2 sentences (before writing the message)
- "What changed?" + "Why?"
- If you cannot describe it cleanly, the commit is probably too big or mixed; go back to step 2.
- Write the commit message
- Use Conventional Commits (required):
type(scope): short summary- blank line
- body (what/why, not implementation diary)
- footer (BREAKING CHANGE) if needed
- Prefer an editor for multi-line messages:
git commit -v - Use
references/commit-message-template.mdif helpful.
- Use Conventional Commits (required):
- Run the smallest relevant verification
- Run the repo's fastest meaningful check (unit tests, lint, or build) before moving on.
- Repeat for the next commit until the working tree is clean
Deliverable
Provide:
- the final commit message(s)
- a short summary per commit (what/why)
- the commands used to stage/review (at minimum:
git diff --cached, plus any tests run)
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
verl-rl-training
Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.
openrlhf-training
High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.
gguf-quantization
GGUF format and llama.cpp quantization for efficient CPU/GPU inference. Use when deploying models on consumer hardware, Apple Silicon, or when needing flexible quantization from 2-8 bit without GPU requirements.
Claude Code Guide
Master guide for using Claude Code effectively. Includes configuration templates, prompting strategies "Thinking" keywords, debugging techniques, and best practices for interacting with the agent.
qdrant-vector-search
High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG systems requiring fast nearest neighbor search, hybrid search with filtering, or scalable vector storage with Rust-powered performance.
behavioral-modes
AI operational modes (brainstorm, implement, debug, review, teach, ship, orchestrate). Use to adapt behavior based on task type.
Didn't find tool you were looking for?