Content
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured, highly actionable skill for a complex multi-mode implementation workflow. Its greatest strengths are the clear step sequencing with explicit gates/checkpoints and the concrete decision trees for handling different modes and edge cases. The main weakness is moderate verbosity — some sections (merge protocol, chain strategy details, workload enforcement) could potentially be extracted to referenced files to keep the main skill leaner, and there's some repetition between steps.
Suggestions
Consider extracting the detailed chain strategy rules (Step 2a) and merge protocol (Step 6) into a referenced file to reduce the main skill's token footprint.
Remove minor redundancies — e.g., 'Mark tasks complete' appears as both part of Step 4's loop and as its own Step 5, which could be consolidated.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is fairly long (~200 lines) with some repetition (e.g., persistence instructions appear in multiple steps, the merge protocol is explained twice). However, most content is genuinely necessary for a complex multi-mode workflow. Some sections like the orchestrator gate preamble and mode explanations could be tighter, but it doesn't over-explain basic concepts. | 2 / 3 |
Actionability | The skill provides highly concrete, executable guidance: specific mem_search queries, exact markdown checkbox syntax, precise decision trees with clear branches, a complete return summary template with all fields, and specific file paths. The pseudocode-style FOR EACH TASK block and decision trees are appropriate for workflow orchestration rather than code generation. | 3 / 3 |
Workflow Clarity | The workflow is clearly sequenced (Steps 1-7) with explicit validation checkpoints: Step 2a enforces a hard gate on workload decisions before any code is written, Step 2b requires reading previous progress before starting, Step 3 has a hard gate for TDD mode, and Step 6 is marked MANDATORY with a merge protocol. Feedback loops are present (blocked states, error reporting, deviation tracking). | 3 / 3 |
Progressive Disclosure | The skill references external files well (sdd-phase-common.md Sections A-D, strict-tdd.md, openspec-convention.md, config.yaml) and the strict-tdd.md module is conditionally loaded only when needed. However, no bundle files were provided to verify these references exist, and some content that could be in referenced files (like the full return summary template and the detailed chain strategy rules) is inline, making the main file longer than necessary. | 2 / 3 |
Total | 10 / 12 Passed |