agentleFS
Sign inStart free
thrillmadethrillmade/agent-skills/skills/conventions-template/SKILL.md

conventions-template

Skill3 starsChanged 8 days agoMITView on GitHub

Template for a "review like me" skill - captures one maintainer's review conventions (formatting, what to flag, what to ignore). Copy + customize per SPEC v0.5.1 applies_to.author.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    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:

    Test coverage on trivial internal helpers

    line 50
  2. Distinctive:

    Prefer pure functions over class hierarchies

    line 57
  3. Distinctive:

    Always cite the specific line + quote the offending snippet

    line 68
  4. Distinctive:

    Always include a concrete suggested fix, not just "consider X"

    line 69
  5. Distinctive:

    Never bundle multiple unrelated issues into one finding

    line 70

≈ “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

Conventions: <YOUR NAME>

This skill captures how I review pull requests when they're opened by me. The bot loads this skill ONLY on my PRs (per the applies_to.author filter above — SPEC v0.5.1 §1.10.1). Other PRs go through the org defaults without my personal preferences applied.

How to use this template

  1. Rename the directory: .claude/skills/conventions-<your-gh-login>/SKILL.md
  2. Set applies_to.author to your GitHub login (replace _REPLACE_ME_). The placeholder fails the SPEC slug check on purpose — if you forget this step, the skill won't load and the bot's log will tell you why.
  3. Fill in the sections below with YOUR actual preferences
  4. Commit + push — the bot picks up the skill on its next review of your PR

The bot still applies the org-wide review-discipline skills (critical-issues-only, evidence-based-review, etc.) on top — your conventions skill LAYERS on, not replaces.

What I always flag

Replace this section with what YOU want the bot to flag on YOUR PRs. Examples:

  • Functions over ~50 lines — split or justify
  • Missing error handling on async/await
  • New dependencies without a one-line "why" in the PR description
  • Magic numbers without a named constant
  • any type in TypeScript (use unknown + narrow)

What I always ignore

What the bot should NOT flag on your PRs (overrides any default behavior that bothers you). Examples:

  • Formatting nits — Prettier handles those
  • JSDoc requirements — I don't use JSDoc, don't ask for it
  • Test coverage on trivial internal helpers
  • "Consider using a more functional approach" suggestions

Structural preferences

Higher-level patterns the bot should respect. Examples:

  • Prefer pure functions over class hierarchies
  • Co-locate tests with source (NOT __tests__/ dirs)
  • One concept per file; files <300 LOC
  • Comments explain WHY, not WHAT (skip comments that restate the code)

How I want findings formatted

Structural rules for how the bot composes findings on your PRs. (Tone — the prose register the bot uses when writing — belongs in a separate voice-<your-gh-login> skill, not here.) Examples:

  • Always cite the specific line + quote the offending snippet
  • Always include a concrete suggested fix, not just "consider X"
  • Never bundle multiple unrelated issues into one finding
  • One finding per file region; merge near-identical findings

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.

Check your setup, free