agentleFS
Diviner-DojoDiviner-Dojo/agent_framework_template/.claude/skills/handling-micro-fixes/SKILL.md

handling-micro-fixes

Skill4 starsChanged 3 months agoApache-2.0View on GitHub

Sizing heuristic and workflow for cosmetic, presentation-only changes (icon/text/label/color/spacing edits). Use when deciding whether a small change is a micro-fix (skip /plan and /build_module) or a behavior change needing the full workflow. Includes the two-strike escalation rule.

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 8 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ask: "Will changing this one thing change behavior?"

    line 31
  2. Distinctive:

    If no → micro-fix. Make the change, run the linter, verify visually, done.

    line 33
  3. Very specific:

    If yes or maybe → not a micro-fix.

    line 34
  4. Very specific:

    Run linter (ruff check / ruff format)

    line 39
  5. Distinctive:

    Verify the result (visually or via test)

    line 40
  6. Distinctive:

    If a micro-fix doesn't produce the expected result after two attempts, STOP coding and switch to diagnosis:

    line 47
  7. Distinctive:

    Add logging/debug output to understand the current state

    line 49
  8. Distinctive:

    Read the relevant code to understand why the change isn't working

    line 50
  9. Distinctive:

    If the root cause is not cosmetic, escalate to the standard workflow

    line 51
  10. Distinctive:

    Rationale: If a "simple" change fails twice, it's not simple.

    line 53

≈ “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

Micro-Fix Protocol

Prevents over-engineering of cosmetic changes. Ensures small fixes stay small.

What Is a Micro-Fix

A change that modifies only the presentation of existing behavior, not the behavior itself:

  • Icon, text, or label changes
  • Color, spacing, or visibility adjustments
  • Menu item reordering
  • CSS/style-only changes
  • Static string corrections

What Is NOT a Micro-Fix

Any change that could alter program behavior, even subtly:

  • Platform API calls, lifecycle callbacks
  • Data pipelines or database schemas
  • Cross-component communication
  • State management changes
  • Network requests or external integrations
  • Configuration that affects runtime behavior

Sizing Heuristic

Ask: "Will changing this one thing change behavior?"

  • If no → micro-fix. Make the change, run the linter, verify visually, done.
  • If yes or maybe → not a micro-fix. Use the standard workflow (/plan if multi-file, /build_module if new module).

Micro-Fix Workflow

  1. Make the change (single file, single concern)
  2. Run linter (ruff check / ruff format)
  3. Verify the result (visually or via test)
  4. Done — no /plan, no /build_module needed

Micro-fixes are NOT exempt from /review or the quality gate when committing. They only skip the planning and build phases.

Two-Strike Escalation Rule

If a micro-fix doesn't produce the expected result after two attempts, STOP coding and switch to diagnosis:

  1. Add logging/debug output to understand the current state
  2. Read the relevant code to understand why the change isn't working
  3. If the root cause is not cosmetic, escalate to the standard workflow

Rationale: If a "simple" change fails twice, it's not simple. Continuing to guess wastes time and risks introducing bugs.

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.