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.
A well-structured, actionable instruction body: numbered steps with concrete CLI commands, explicit stop points, and error-recovery guardrails. The only notable nit is minor repetition of the --schema guidance.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and instruction-dense with no explanations of concepts Claude already knows; the only fat is the --schema rule repeated three times (step 2, step 3, and guardrails). Not 5 because that redundancy could be trimmed to a single statement. | 4 / 5 |
Actionability | Gives copy-paste-ready commands (`openspec new change "<name>"`, `openspec status --change "<name>"`, `openspec instructions <first-artifact-id> --change "<name>"`) and a concrete AskUserQuestion prompt. Not 5 because <first-artifact-id> depends on interpreting status output and no CLI-failure handling is specified. | 4 / 5 |
Workflow Clarity | Six explicitly numbered steps with a hard STOP checkpoint (step 6), a status-verification step, and error-recovery guardrails for the main failure modes (invalid kebab-case name, duplicate change name). No destructive or batch operations, so no validation cap applies. | 5 / 5 |
Progressive Disclosure | Self-contained single-file skill with well-organized sections (Input, Steps, Output, Guardrails), no bundle files, and no content that belongs in a separate reference — meets the simple-skill exception for well-organized sections. | 5 / 5 |
Total | 18 / 20 Passed |