Content
67%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, highly actionable instruction-only skill with a clear ordered workflow, explicit validation checkpoints, and a properly wired one-level-deep reference bundle. Its weaknesses are redundancy across the caution sections and overlap between inline output specs and the reference modules, plus one missing worked example.
Suggestions
Consolidate the overlapping caution sections ("Core Function should not", "Hard Rules" 3/7/13/14, "What This Skill Should Not Do", "Quality Standard") into a single rules section to cut repetition and tokens.
Remove the inline Section A-I column-level specifications that duplicate references/output-section-guidance.md, keeping only section-level requirements in the body and delegating formatting detail to the reference.
Add a short worked example of a filled Section C opportunity-map row (one path, one bridge-evidence label) so the required table format is unambiguous.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient and free of filler, but the same cautions are restated across four overlapping sections ("Core Function should not", Hard Rules 3/7/13/14, "What This Skill Should Not Do", and "Quality Standard"), e.g. never treating mechanism relevance as translational usability appears in all of them. Consolidating these would tighten the file without losing information. | 3 / 5 |
Actionability | The 8-step execution, per-path audit checklists, per-section table column specs (Sections C-F), and verbatim templates (redirect message, check-in prompt, Limited Evidence Mode flag, retrieval fallback label) give directly executable guidance. Minor gaps: no worked example of a filled opportunity table, and Step 2 gives retrieval priorities but no concrete search strategy. | 4 / 5 |
Workflow Clarity | The "8 Steps (always run in order)" sequence includes explicit checkpoints (Step 1.5 scope check-in, Step 8 pre-finalize self-review, Section I risk review, retrieval fallback) plus a 15-rule hard checklist. It falls short of 5 because there is no explicit error-recovery loop stating what to revise and re-check if Step 8's review finds a problem. | 4 / 5 |
Progressive Disclosure | All seven references/ files exist, are clearly signaled in the Reference Module Integration section with when-to-use mapping, and are one level deep with no nested references. However, the body duplicates reference content inline (Sections A-I column specs overlap output-section-guidance.md, and Step 3's path list overlaps the use-case framework), which keeps it below the top anchor. | 4 / 5 |
Total | 15 / 20 Passed |