OpenParable
Skip to content
← Back to skills

spool-commit

Development

Create atomic git commits aligned to Spool changes. Use when you want to commit work after applying a change, optionally with auto-mode.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/workflow/spool-commit/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/spool-commit, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Create atomic git commits aligned to Spool changes.

Concept: In Spool-driven workflows, you typically make progress by creating/applying a change. After applying and verifying a change, you should usually create a git commit that corresponds to that change.

Key behavior

  • Prefer 1 commit per applied Spool change (or a small number of commits if the change is large).
  • Include the Spool change id in the commit message when practical (e.g. 001-02_add-tasks).
  • Use Spool inspection commands to anchor the commit to what was actually applied.

Parameters

When invoking this skill, check for these parameters in context:

  • auto_mode: boolean flag

    • true: create commits immediately without asking for confirmation
    • false or missing: ask for confirmation of each commit message
    • CRITICAL: this only applies to the current invocation and is reset afterwards

  • change_id: optional, a Spool change id (recommended)

    • If missing, prompt the user to pick from spool list --json

  • stacked_mode: optional boolean

    • If true, create stacked branches per commit (only if tooling exists)
    • If false or missing, commit on current branch

  • ticket_id: optional identifier to include in commit messages

Prerequisites

  1. Verify repo has changes:

    • git status --short
    • If no changes, stop with: "No changes found to commit"

  2. Identify Spool change context:

    • If change_id not provided, run spool list --json and ask user to select
    • Then inspect the change: spool status --change "<change-id>"

  3. Confirm the change is in a good commit state:

    • Ensure artifacts/tasks are complete enough that a commit makes sense
    • If the change is unfinished, ask user whether to commit "WIP" or wait

Commit Message Format

Use conventional commit format:

  • Format: type(scope): description
  • Prefer scope = Spool module name or ticket id
  • Description should mention the change goal
  • Include Spool change id at end, in parentheses, when practical

Examples:

  • feat(todo): add task model and parsing (001-02_add-task-core)
  • fix(storage): persist tasks atomically (002-01_storage-save)

Procedure

  1. Read diffs: git diff and git status --short

  2. Stage files for the selected change (prefer staging only files touched by that change)

  3. Decide commit messages:

    • If auto_mode is true: commit immediately
    • Otherwise: present recommended message + alternatives, ask user to confirm

  4. Verify after each commit: git status --short

Output

After committing, show:

  • Change committed:
  • Commit SHA + message (git log -1 --oneline)
  • Remaining uncommitted changes (if any)