commit
Draft a Conventional Commit message in English, get explicit user approval, then commit; never bypasses hooks.
Skill by ccosming
Chalk Labs
Need help deploying commit?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/ccosming-commit
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: writing
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/commit
Tools it pre-approves 4
From allowed-tools in SKILL.md: the agent may use these without asking you each time.
Bash,Read,Grep,AskUserQuestion
What's inside
- Skill name
commit- Description
- Draft a Conventional Commit message in English, get explicit user approval, then commit; never bypasses hooks. Use when asked to "commit", "make a commit", "stage and commit", or "write a commit message".
- License
- MIT
- Allowed tools
- Bash,, Read,, Grep,, AskUserQuestion
- Instructions
- 150 lines in SKILL.md · view on GitHub
Before installing
- Pre-approves some shell commands
- Instructions read files in your home folder (~/.ssh, ~/.aws…)
Files (1)
- SKILL.md
SKILL.md (first lines)
# Git commit with approval Creates a commit following [Conventional Commits](https://www.conventionalcommits.org/). Message **in English**, no `Co-Authored-By` line, no `--no-verify`, no `--amend` to "rescue" a failed commit. ## Workflow ### 1. Inspect state Run these in **parallel** (single assistant turn, multiple `Bash` tool calls): ```bash git status git diff --stat git diff --staged --stat git log -5 --pretty=format:'%h %s%n%b%n---' ``` **Abort conditions** — report and stop, do not continue: - Working tree clean and nothing staged → "No changes to commit." - Files in conflict (`UU` rows in `git status`) → "Conflicts present; resolve before committing." - Nothing staged but unstaged changes exist → list the unstaged files and ask the user which ones to stage before continuing. ### 2. Detect potential secrets Before drafting, audit the staged surface: | Check | How | |---|---| | Sensitive filenames | Inspect staged file list for: `*.env`, `*.pem`, `*.key`, `id_rsa*`, `credentials*`, `*secret*`, `*token*` | | Sensitive content | Scan `git diff --staged` output for: `AWS_SECRET`, `API_KEY=`, `-----BEGIN .* PRIVATE KEY-----`, `password=`, long opaque base64/hex tokens | If any hit, **pause**: surface the finding to the user in the proposal turn (step 4) and require explicit acknowledgement before executing. ### 3. Draft the message **Caller context (e.g. from `/code`):** when another skill invokes this one with a suggested scope/subject in its args (e.g. `suggested_scope: FEAT-007; subject: pure validation logic`), seed the draft with it. Steps 1, 2, and 4 still run in full — the suggestion never skips the secrets check or the approval gate. Direct user invocations carry no such args; draft from the diff. Determine: - **Type** — one of: `feat`, `fix`, `refactor`, `docs`, `chore`, `test`, `style`, `perf`, `ci`, `build`. - **Scope** (optional) — affected area, e.g. `feat(extensions)`, `fix(commit)`, `chore(toolchain)`. Pick the narrowest accurate scope. - **Subject** — imperative mood, lowercase after `type(scope):`, ≤ 72 chars, no trailing period. - **Body** — atomic bullets prefixed with `-`. One change per bullet. No prose narrative, no "this commit…" framing. - **No internal references** — the message describes the change on its own terms. Never cite internal process labels (phase names, finding/resolution IDs) or gitignored working notes (e.g. `notes/`); those are private to the workflow, not part of the repository. Format: ``` type(scope): concise subject line - bullet capturing change 1 - bullet capturing change 2 ``` If the change set spans unrelated concerns, **propose splitting** into multiple commits instead of one fat message. ### 4. Propose and confirm via AskUserQuestion Present the proposal as text: ``` ## Proposed commit **Files**: - path/one - path/two **Message**: type(scope): subject - bullet 1 - bullet 2 **Secrets check**: clean | <list findings> ``` Then collect approval with **`AskUserQuestion`** — never open free-text: - One question, options **Approve** | **Adjust**. - The user's "Other" free-text or **Adjust** = a change request: revise the draft and repeat step 4. - Commit only when the user picks **Approve** → step 5. - If a secrets finding surfaced, state it in the proposal and require an explicit Approve that acknowledges it before executing. ### 5. Execute 1. Stage **only** the files listed in the proposal: ```bash git add path/one path/two ``` **Never** use `git add -A`, `git add .`, or wildcards beyond what was approved. 2. Commit using a HEREDOC to preserve newlines: ```bash git commit -m "$(cat <<'EOF' type(scope): subject - bullet 1 - bullet 2 EOF )" ``` 3. Verify and report: ```bash git status git log -1 --pretty=format:'%h %s' ``` Output the resulting commit hash to the user.
From ccosming/skills, MIT.
More from ccosming/skills
- 0.479score
humanizer
ccosming
Strip AI-writing tells from text to make it sound natural and human.
SkillCFreshness: CClaude Code +9 - 0.474score
coding-rules
ccosming
Behavioral guidelines to reduce common LLM coding mistakes.
SkillCFreshness: CClaude Code +9 - 0.474score
create-skill
ccosming
Interactive guide for creating a new Claude Code skill in this repository.
SkillCFreshness: CClaude Code +9 - 0.469score
ideate
ccosming
Turns a half-formed idea into a standalone concept whitepaper, interviewing one question at a time.
SkillCFreshness: CClaude Code - 0.469score
categorizer
ccosming
Internal classifier of the spec workflow.
SkillCFreshness: CClaude Code +9 - 0.469score
charter
ccosming
Internal charter-authoring stage of the spec workflow.
SkillCFreshness: CClaude Code - 0.469score
consistency
ccosming
Reads one assembled `.spec/` artifact cold and reports where it breaks: sections that contradict each other (a section depending on something another…
SkillCFreshness: CClaude Code +9 - 0.469score
detector
ccosming
Reads a just-finished authoring exchange and the artifact it produced, and reports cross-artifact material that should seed or update another artifact — a…
SkillCFreshness: CClaude Code +9 - 0.469score
drafting
ccosming
Internal drafter of the spec workflow.
SkillCFreshness: CClaude Code +9 - 0.469score
grill
ccosming
Interviews the user one question at a time about a half-formed idea until a shared understanding emerges.
SkillCFreshness: CClaude Code +9 - 0.469score
arch
ccosming
Internal architecture-authoring stage of the spec workflow.
SkillCFreshness: CClaude Code - 0.467score
adr
ccosming
Internal ADR-authoring stage of the spec workflow.
SkillCFreshness: CClaude Code
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Jun 17, 2026 (4 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/ccosming/skills
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from ccosming
- 0.479score
humanizer
ccosming
Strip AI-writing tells from text to make it sound natural and human.
SkillCFreshness: CClaude Code +9 - 0.474score
create-skill
ccosming
Interactive guide for creating a new Claude Code skill in this repository.
SkillCFreshness: CClaude Code +9 - 0.474score
coding-rules
ccosming
Behavioral guidelines to reduce common LLM coding mistakes.
SkillCFreshness: CClaude Code +9 - 0.469score
consistency
ccosming
Reads one assembled `.spec/` artifact cold and reports where it breaks: sections that contradict each other (a section depending on something another…
SkillCFreshness: CClaude Code +9 - 0.469score
charter
ccosming
Internal charter-authoring stage of the spec workflow.
SkillCFreshness: CClaude Code - 0.469score
arch
ccosming
Internal architecture-authoring stage of the spec workflow.
SkillCFreshness: CClaude Code
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get commit running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
commit skill. Agent Mart. https://agentmart.dev/skills/ccosming-commit (last updated 2026-10-24)