agentleFS
theseus-runtheseus-run/theseus/.agents/skills/biome-grit-rules/SKILL.md

biome-grit-rules

Skill1 starsChanged 6 months agoMITView on GitHub

Use when creating, reviewing, or modifying custom Biome Grit plugins in plugins/*.grit, especially deterministic agent-correction rules for stale Effect v3 patterns, local style hazards, and diagnostic messages that tell agents what to do next.

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. About 0 are common, 9 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

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

    Use this skill for repo custom lint rules under plugins/*.grit.

    line 8
  2. Distinctive:

    Add or tighten a rule only when:

    line 16
  3. Distinctive:

    Every diagnostic message should act as a small prompt:

    line 28
  4. Distinctive:

    Do not hard-code exact dependency versions unless the version itself is the rule.

    line 41
  5. Very specific:

    Use error for stale APIs, known unsafe patterns, and hard repo invariants.

    line 45
  6. Very specific:

    Use warning for style guidance or patterns with legitimate exceptions.

    line 46
  7. Distinctive:

    Avoid broad rules that force taste when the repo does not have a stable invariant.

    line 47
  8. Distinctive:

    After changing a rule:

    line 51
  9. Very specific:

    Run bunx biome check <target> against a file that should trigger or currently triggered the rule.

    line 53
  10. Distinctive:

    Run the focused check against at least one nearby clean file when false positives are plausible.

    line 55
  11. Very specific:

    Run bun run lint when the rule is broad or touches shared config.

    line 56
  12. Very specific:

    Keep plugin filenames specific and negative: no-effect-v3-catchall.grit.

    line 60
  13. Distinctive:

    Keep messages short enough to read in terminal output.

    line 61
  14. Distinctive:

    Prefer rules that remove a whole class of future fixes.

    line 62
  15. Distinctive:

    If a rule fires too broadly, narrow the pattern before lowering severity.

    line 63

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

Biome Grit Rules

Use this skill for repo custom lint rules under plugins/*.grit.

Purpose

Custom rules turn repeated agent mistakes into deterministic feedback. In this repo, Effect v3 patterns are a primary target because they are common in model training data while Theseus uses Effect v4 patterns.

Before Adding A Rule

Add or tighten a rule only when:

  • the bad pattern is syntactically recognizable
  • the desired replacement is deterministic or clearly explainable
  • false positives are low
  • the pattern has appeared more than once, is high-risk, or reflects stale API training data
  • the rule can produce a useful diagnostic at the exact bad span

Propose first instead of editing when the rule could affect many files or needs semantic judgment.

Diagnostic Messages

Every diagnostic message should act as a small prompt:

  • name the repo rule or API generation, such as Effect v4
  • say what is wrong
  • say what to do instead
  • mention the relevant skill when the mistake implies a stale mental model

Example shape:

Effect v4: Effect.catchAll was renamed to Effect.catch. Use Effect.catch for all typed errors, Effect.catchTag/Effect.catchTags for specific tags, or Effect.catchCause for cause-level handling. Load skill: effect-v4.

Do not hard-code exact dependency versions unless the version itself is the rule.

Severity

  • Use error for stale APIs, known unsafe patterns, and hard repo invariants.
  • Use warning for style guidance or patterns with legitimate exceptions.
  • Avoid broad rules that force taste when the repo does not have a stable invariant.

Verification

After changing a rule:

  1. Run bunx biome check <target> against a file that should trigger or currently triggered the rule.
  2. Confirm the diagnostic span and message are clear.
  3. Run the focused check against at least one nearby clean file when false positives are plausible.
  4. Run bun run lint when the rule is broad or touches shared config.

Maintenance

  • Keep plugin filenames specific and negative: no-effect-v3-catchall.grit.
  • Keep messages short enough to read in terminal output.
  • Prefer rules that remove a whole class of future fixes.
  • If a rule fires too broadly, narrow the pattern before lowering severity.

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.