Content
85%Weight 40%Scale 1-5Reviews 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, highly actionable contract with a clear six-step workflow, explicit verification, and proper one-level-deep delegation of the ADR schema to a real reference file. Its only weakness is redundant restatement of the same boundary rules across the Input, Qualification handoff, Workflow, and Boundaries sections.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is imperative and free of concept explanations Claude already knows, but the same constraints are restated across sections (one-decision rule in Input, step 1, and Boundaries; gate-ownership in the handoff section and step 1; reporting ownership in Purpose, step 4, and step 6), so it could be tightened beyond minor trimming. | 3 / 5 |
Actionability | Fully concrete instruction-only guidance: exact path pattern `context/decisions/YYYY-MM-DD-<decision-slug>.md`, the exact allowed status set, exact result fields per outcome, and a real template file (references/adr-template.md) delegated as the sole schema authority. | 5 / 5 |
Workflow Clarity | Six numbered steps with an explicit validation checkpoint (step 5 verifies creation/reuse count, schema conformance, referenced paths, and that no existing ADR was modified) and explicit error-recovery paths (`blocked` with problem/impact/required_action; deterministic slug collision resolution). | 5 / 5 |
Progressive Disclosure | The body is an overview/contract and the detailed ADR schema is appropriately split into references/adr-template.md, which exists, is clearly signaled ("the sole authority for the persisted ADR schema"), and is one level deep with no nested references. | 5 / 5 |
Total | 18 / 20 Passed |