agentleFS
woodbearywoodbeary/later-dog/.agents/skills/correct-recurring-errors/SKILL.md

correct-recurring-errors

Skill2 starsChanged 29 days agoApache-2.0View on GitHub

When the same failure has happened more than once and the evidence shows it, prevent it at the narrowest lasting place: a type or structural fix, a CI check or fixture, or a short rule in the canonical instructions.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    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:

    Only if neither works, write a short rule where agents will read it: AGENTS.md, or the canonical skill in skills/laterdog/ followed by pnpm …

    line 17
  2. Distinctive:

    Show it works: Run the prevention against the failing case (revert the fix or replay the input) and watch it fail, then against a good case …

    line 21

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 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

Correct recurring errors

Use when a mistake comes back: the same review finding, the same failed check, the same wrong assumption in a job brief, the same correction from the person. Once is a bug to fix. Twice is a pattern to prevent. Use the later.dog MCP server. Discover tools by the names below; client-specific prefixes may vary.

Procedure

  1. Gather the cases. Collect every occurrence with its evidence: job receipts from inspect_cloud_job, verification and review results, repair runs, the person's corrections in chat, CI logs. With fewer than two evidenced cases, fix the bug and stop.
  2. Name the cause in one sentence: the condition that triggers it and what goes wrong. Cases without a shared cause are separate bugs.
  3. Pick the strongest prevention that fits, in this order:

    1. Make it impossible: a type, a schema, a single source of truth, a narrower interface.
    2. Make CI catch it: a lint rule, a test, a fixture or a verification recipe (docs/verification/, docs/laterdog/verification.md).
    3. Only if neither works, write a short rule where agents will read it: AGENTS.md, or the canonical skill in skills/laterdog/ followed by pnpm laterdog:skills. Never edit an exported copy.

    Counting on someone to spot it in review is not a prevention.

  4. Keep it narrow. The prevention names the trigger and the failure. It does not fence off a whole area because one path through it went wrong.
  5. Show it works. Run the prevention against the failing case (revert the fix or replay the input) and watch it fail, then against a good case and watch it pass.

Report

The cases, the cause, the prevention and why the stronger options did not fit, what you ran to prove it, and any provider or environment limit that remains.

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.