Content
86%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.
A highly actionable, well-sequenced CLI-driven workflow with exact commands, response fields, and clear branching per status. The only weaknesses are mild repetition between the steps and guardrails and the absence of an error-recovery loop.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is command-first and free of explanations Claude already knows, with tight bullets and exact JSON field names. Minor duplication could be trimmed: the 'do NOT include context/rules in output' rule appears in Step 3 and again in Guardrails, and 'STOP after creating ONE artifact' repeats the one-artifact guardrail. | 4 / 5 |
Actionability | Three copy-paste-ready CLI commands with exact flags ('openspec list --json', 'openspec status --change "<name>" --json', 'openspec instructions <artifact-id> --change "<name>" --json'), exact response fields to parse, and explicit per-status branching cover the common cases, matching the fully-executable anchor. | 5 / 5 |
Workflow Clarity | A clear four-step sequence with explicit branches (complete / ready / all blocked) and a verification guardrail ('Verify the artifact file exists after writing before marking progress'). It lacks a validate-fix-retry feedback loop — the 'all blocked' branch only says to 'suggest checking for issues' — so it stops short of the level-5 anchor; the single-artifact, non-destructive nature means no cap applies. | 4 / 5 |
Progressive Disclosure | No bundle files exist and none are referenced; the body is well-sectioned (Input, Steps, Output, Artifact Creation Guidelines, Guardrails) with no inlined reference material and no nested references. Schema-specific detail is delegated to the CLI at runtime, which is an appropriate progressive-disclosure mechanism. | 5 / 5 |
Total | 18 / 20 Passed |