Content
100%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A tight, expert-aimed decision layer that routes test work to the right tier with concrete commands, repo-specific traps, and a well-signaled companion reference. It assumes Claude's competence and concentrates entirely on non-obvious project knowledge.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and dense with repo-specific knowledge Claude lacks (tier routing, traps, exact paths); it never explains basic concepts like what a unit test or Postgres is, so every token earns its place. | 3 / 3 |
Actionability | Provides concrete executable commands and paths — 'cargo test --features integration', 'grep -n integration .github/workflows/platform-and-compat.yml', 'bash scripts/reborn-e2e-rust.sh' — alongside specific file targets per tier. | 3 / 3 |
Workflow Clarity | The numbered tier decision tree and sequenced Verify pipeline give a clear routing sequence, with explicit re-verify checkpoints ('re-verify: grep …', 'run it locally when your change is DB/runtime-shaped') acting as validation gates. | 3 / 3 |
Progressive Disclosure | Well-organized sections form a concise overview, and the single one-level-deep reference [references/exemplar-tests.md] is clearly signaled and verified present, keeping detail out of the main file. | 3 / 3 |
Total | 12 / 12 Passed |