Content
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is highly actionable with a clear sequenced workflow, explicit checkpoints, and a ready-to-use template. Its main weaknesses are verbosity in the question-orchestration section and unverifiable external shared-file references that hurt progressive-disclosure structure.
Suggestions
Tighten the 10-item proposal-question enumeration into a compact checklist or move it to a references file to reduce inline token weight.
Either include the referenced 'skills/_shared/sdd-phase-common.md' and 'openspec-convention.md' as bundle files, or inline the minimal Section A/B/C/D content so the skill is self-contained.
Condense the orchestrator-gate prose at the top; the gate mechanism could be stated in fewer lines without the repeated emphasis.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is thorough and mostly earns its tokens, but the 10-item proposal-question enumeration and verbose orchestrator-gate prose could be tightened; some explanation restates concepts an executor already understands. | 2 / 3 |
Actionability | Provides a complete, copy-paste-ready proposal.md template with concrete sections, tables, topic keys, and explicit artifact persistence steps with real keys like 'sdd/{change-name}/proposal'. | 3 / 3 |
Workflow Clarity | Sequenced as numbered Steps 0–6 with explicit mandatory checkpoints ('This step is MANDATORY — do NOT skip it'), mode-branching IF/ELSE conditions, and a validation-like 'READ it first and UPDATE it' guard for existing proposals. | 3 / 3 |
Progressive Disclosure | No bundle files exist (no references/, scripts/, assets/), yet the body references external shared files like 'skills/_shared/sdd-phase-common.md' and 'skills/_shared/openspec-convention.md' that are not present in the bundle, so the one-level-deep reference structure cannot be verified. | 2 / 3 |
Total | 10 / 12 Passed |