Conciseness

The body is lean and imperative throughout — no padding, no explanation of concepts an agent already knows (e.g., 'Prefer ranges, syntax nodes, tokens, and compact enums in State', 'Set FixKind::Safe only when no reachable input changes behavior'). Every line prescribes or routes to a specific resource. The Review Checklist is verification content, not filler, and each token earns its place.

Actionability

Fully concrete and executable: exact scaffolding commands ('just new-js-lintrule useMyRule', 'just new-js-assistrule useMyAction'), test commands ('just test-lintrule useMyRule', 'just gen-rules', 'just gen-configuration', 'just f', 'just l'), specific API surface ('Ast<Node>', 'Semantic<Node>', 'FixKind::Safe', 'run()' vs 'action()'), and exact repo paths for rules, options, and fixtures.

Workflow Clarity

A clear six-step workflow with embedded validation — 'Add focused valid and invalid fixtures, then inspect snapshots' and 'Run required analyzer codegen and the narrowest tests' — plus explicit action-verification requirements ('Test that applying the action: produces valid syntax; preserves comments and trivia...') and a dedicated Review Checklist for error recovery before finishing.

Progressive Disclosure

The body is a well-organized overview that routes detail one level deep with clearly signaled references: 'Follow the current analyzer architecture in crates/biome_analyze/CONTRIBUTING.md. Read only the sections relevant to the rule', per-topic 'Load X for Y' cross-skill pointers (diagnostics-development, testing-codegen, changeset, syntax-text-handling, eslint-migrate-options), and a closing References section of exact repo paths. No nested references, no bulk detail inlined.

Specificity

The description enumerates concrete, comprehensive coverage of the skill's scope: 'creating or modifying Biome lint rules or assists, including analyzer queries, semantic bindings, rule state, code actions, fix safety, options, registration, and end-user rule rustdoc'. Every listed item is a specific capability area, not generic filler, and coverage spans the full rule-development lifecycle.

Completeness

It explicitly answers both questions: 'Use this skill when creating or modifying Biome lint rules or assists' is a clear 'when' trigger, and the enumerated scope (queries through rustdoc) is a clear 'what'. It even adds routing guidance for adjacent skills ('Also load diagnostics-development... and testing-codegen...').

Trigger Term Quality

Good keyword coverage with natural terms a Biome contributor would say: 'lint rules', 'assists', 'analyzer', 'code actions', 'fixtures', 'snapshot', 'rustdoc', 'options'. A few natural synonyms and variations are missing (e.g., 'linter', 'diagnostic', 'suppression', 'plugin'), so it falls just short of the comprehensive-synonyms anchor.

Distinctiveness Conflict Risk

It occupies a clear niche — Biome lint rule/assist development — with domain-specific triggers ('Biome', 'analyzer queries', 'nursery'-adjacent terminology) that would not conflict with general linting, diagnostics, or testing skills for other tooling.