Content
81%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, largely executable workflow with strong sequencing and genuine validation checkpoints (status checks, confirmation prompts, target-collision handling) before the destructive archive move. Weaknesses are minor: some redundancy between the Guardrails section and the steps, a somewhat high-level delta-spec comparison step, and no offloading of the sync-assessment detail despite the body's length.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body gives commands, paths, and decision rules directly with no explanations of concepts Claude already knows, e.g. 'Run `openspec status --change "<name>" --json`' and 'Count tasks marked with `- [ ]`'. It is not a 5 because the Guardrails section repeats steps 1, 4, and 6 ('Always prompt for change selection if not provided', 'Show clear summary of what happened'), and the opening line duplicates the description. | 4 / 5 |
Actionability | Most steps are executable with copy-paste commands ('mkdir -p openspec/changes/archive', 'mv openspec/changes/<name> openspec/changes/archive/YYYY-MM-DD-<name>') and a concrete Task-tool invocation including subagent_type and prompt template. It falls short of anchor 5 because step 4's 'Determine what changes would be applied' is high-level with no command or method for comparing delta specs to main specs. | 4 / 5 |
Workflow Clarity | Six clearly sequenced steps with explicit validation checkpoints before the destructive move: parsing artifact status, counting incomplete tasks, checking 'If target already exists: If yes: Fail with error, suggest renaming existing archive or using different date', plus AskUserQuestion confirmation loops for incomplete artifacts/tasks. This matches anchor 5 (explicit validation, feedback loops for error recovery), and the destructive-operation cap does not apply since validation is present throughout. | 5 / 5 |
Progressive Disclosure | The body is well organized into Steps, Output On Success, and Guardrails sections with no nested references and only a one-level reference to the openspec-sync-specs skill. Not a 5 because at roughly 93 body lines it exceeds the simple-skill (<50 line) allowance for a top score without external references, and the delta-spec sync assessment details could live in a reference file. | 4 / 5 |
Total | 17 / 20 Passed |