agentleFS
hellocrossmanwildhellocrossmanwild/Claude-skills-tools/skills/safe-change/SKILL.md

safe-change

Skill0 starsChanged 38 days agoMITView on GitHub

Makes code changes safely: a branch, plain-English save points, a pull request and easy undo. Use for changes to a live app, or 'undo that', 'go back', 'what changed' or 'open a PR'.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 6 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If the project is already live, make a branch for this change (name it after the change, e.g.

    line 12
  2. Distinctive:

    Check there are no unsaved changes lying around.

    line 13
  3. Distinctive:

    Keep each change small and about one thing.

    line 16
  4. Distinctive:

    After each piece works, make a save point (commit) with a note a non-developer would understand: "Add a reminder email the day before a …

    line 17
  5. Distinctive:

    If they say "undo", "go back" or "that broke it":

    line 29
  6. Distinctive:

    Ask which one to go back to.

    line 31
  7. Distinctive:

    If they ask "what changed?" or "explain this", describe the change in plain English: which screens or behaviour it affects, not file names …

    line 36

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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

Safe change

You're making changes for someone who doesn't read code. Your job is to keep the live version safe, explain every change in plain English, and make undoing easy. Use the GitHub connector where it helps.

Before changing anything

  1. Say in one or two plain sentences what you're about to change and why.
  2. If the project is already live, make a branch for this change (name it after the change, e.g. add-booking-reminders). Never work straight on main once real people use the app. For a brand-new project's first build, working on main is fine.
  3. Check there are no unsaved changes lying around. If there are, ask what to do with them before starting.

While changing

  • Keep each change small and about one thing.
  • After each piece works, make a save point (commit) with a note a non-developer would understand: "Add a reminder email the day before a booking", not "refactor scheduler".
  • Push the branch to GitHub so the work is backed up.

When it's ready

Open a pull request with:

  • What changed: plain-English bullet points
  • Why: one sentence
  • How to check it: the exact steps to try it (and the preview link if Vercel made one)
  • Anything risky: data changes, new keys needed, things that can't easily be undone Then stop and let them review. Don't merge unless they say so.

Undo

If they say "undo", "go back" or "that broke it":

  1. List the recent save points in plain English, newest first.
  2. Ask which one to go back to.
  3. Go back without losing history: prefer reverting (a new save point that undoes the change) over deleting history. Never force-push or delete branches without a clear yes.
  4. Confirm what's now live and what was undone.

Explaining a change

If they ask "what changed?" or "explain this", describe the change in plain English: which screens or behaviour it affects, not file names or code.

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.