Content
100%Weight 40%Scale 1-3Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a well-structured, actionable catalog of Sentry frontend bug patterns with concrete safe-pattern code, a clear gated workflow, and correctly signaled one-level-deep references. It stays lean by assuming Claude's competence and avoiding concept explanations.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense with domain-specific, non-obvious guidance (red flags plus concrete safe-pattern code) and assumes Claude's competence without explaining basics like React hooks or null access; the only mild excess is a one-sentence motivational framing of the production-issue statistics. | 3 / 3 |
Actionability | Each check pairs red flags with copy-paste-ready safe patterns such as `obj?.property?.nested`, `const value = obj?.field ?? defaultValue`, `if (!traceSlug) return fallbackLink`, and `Math.min(100, Math.max(0, precision))` — fully executable guidance. | 3 / 3 |
Workflow Clarity | A clearly sequenced three-step process (Classify → Check patterns → Report) with explicit checkpoints: a HIGH/MEDIUM/LOW confidence gating table, the 'If no checks produced a potential finding, stop and report zero findings' stop rule, and a dedup rule for reporting each location once. | 3 / 3 |
Progressive Disclosure | The body is an overview that delegates per-category detail to seven one-level-deep references via a clear code-type-to-reference table in Step 1, and all seven referenced files (e.g. references/null-reference-errors.md) exist in ./references/. | 3 / 3 |
Total | 12 / 12 Passed |