agentleFS
ch4570ch4570/vulpora/skills/git-flow/SKILL.md

git-flow

Skill1 starsChanged 56 days agoApache-2.0View on GitHub

Prepare a safe work branch or commit, push, and open a GitLab merge request when the user asks to start, land, or publish implementation work.

We split this skill into the rules it gives an agent

3 rules. 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 3

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before implementation, read branch preparation.

    line 16
  2. Distinctive:

    Keep the current branch; do not run branch preparation over existing changes.

    line 19
  3. Distinctive:

    Read principles only for rationale, and the KB index only to find a needed transport or policy topic.

    line 21

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

git-flow — Work Branch and Commit/Push/MR Skill

Purpose

Prepare an updated child branch before implementation, or land an existing non-protected work branch through explicit staging, Korean commit messages, push, and one GitLab merge request.

Prerequisites

Select the requested path before any Git mutation:

  • Before implementation, read branch preparation. This path ends after the child branch is ready; it does not load commit, transport, or MR recipes.
  • To land existing work, read the complete commit/push/MR contract. Keep the current branch; do not run branch preparation over existing changes.

Read principles only for rationale, and the KB index only to find a needed transport or policy topic. They are not unconditional first reads. Both paths preserve Korean user-facing messages, explicit target/default develop, existing user changes, and the stop conditions below.

Instructions

  1. Bind the explicit target branch, otherwise develop; require a clean, non-diverged base before creating a child branch.
  2. For existing work, stop on protected branches, unsafe staged files, missing authority, dirty base preparation, or unresolvable transport.
  3. Stage explicit in-scope paths, create coherent Korean commits, check divergence before each push, and create or reuse exactly one MR targeting the bound base.
  4. Preserve the contract's transport order, secret guard, stop conditions, MR fields, and post-creation verification. Never force-push, rebase, merge, or auto-stash.

Examples

  • New task: fast-forward develop, create a task-derived child branch, then return to implementation without landing an empty commit.
  • Existing work: inspect the diff, stage only the coherent change set, push after the divergence check, and create or reuse the MR.

Limitations

Code review, CI polling, MR merge, branch deletion, force-push recovery, and rebase are outside this skill. The detailed command recipes and MR template remain in the operating contract to keep normal skill loading small.

Troubleshooting

ConditionResolution
Dirty or diverged basePreserve it and stop; do not stash, reset, or merge around it.
No authenticated GitLab transportGive the contract's recovery guidance and stop unless an approved fallback is available.
Existing MRReuse it and verify its canonical IID and URL.

Verify

Report the branch, transport, MR IID/URL, staged scope, verification evidence, and any remaining risk.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.