Content
75%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 content is well-structured and highly actionable, with concrete commands and validation checkpoints appropriate for a destructive archive operation. It is held back from top marks by minor redundancy, a small gap in date-generation guidance, and lack of a retry loop / post-archive verification.
Suggestions
Add an explicit post-archive verification step (e.g., confirm the archived directory exists and the source is gone) to strengthen the workflow feedback loop.
Provide a concrete command for generating the current date (e.g., date +%Y-%m-%d) instead of the bare YYYY-MM-DD placeholder.
Trim the Guardrails section of items that restate step instructions (e.g. 'Always prompt for change selection if not provided') to reduce redundancy.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is direct and assumes Claude's competence (no explanations of JSON or archiving), with only minor redundancy such as the Guardrails section restating step content like 'Always prompt for change selection if not provided'. | 4 / 5 |
Actionability | Provides concrete, copy-paste-ready commands (openspec list --json, openspec status --change, mkdir -p, mv) and specific JSON fields to parse, with a minor gap around generating the current date for the archive name. | 4 / 5 |
Workflow Clarity | A clear 6-step sequence with explicit validation checkpoints (artifact/task completion checks with AskUserQuestion confirmation, and a pre-move target-exists guard) for a destructive-ish directory move; not a 5 because error recovery is 'fail and suggest' rather than a validate-fix-retry loop with post-archive verification. | 4 / 5 |
Progressive Disclosure | Self-contained skill with no bundle files needed and a clean section structure (Input, Steps, Output On Success, Guardrails); slightly above the simple-skill line and well-organized, though not a 5 because some inline content (e.g. the sync subagent prompt and output template) is dense. | 4 / 5 |
Total | 16 / 20 Passed |