Content
92%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.
An excellent orchestration skill body: highly actionable workflows with hard gates, validation, and retry loops, plus a clean progressive-disclosure structure where every reference resolves to a real file. The only weakness is redundant restatement of a few enforcement rules, which costs tokens without adding clarity.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense operational guidance with no concept-teaching padding, but enforcement rules are repeated (prompt-file-before-backend stated in four places; the confirmation gate restated in Confirmation Policy, Step 2, and Notes) and could be consolidated. | 4 / 5 |
Actionability | Fully concrete for an instruction-only skill: exact file-layout tree, prompt naming convention (NN-{type}-{slug}.md), EXTEND.md search paths, backend resolution priority order, session-ID format (cards-{topic-slug}-{timestamp}), and verbatim watermark prompt text. | 5 / 5 |
Workflow Clarity | The Step 0-4 checklist has explicit gates (Step 0 BLOCKING, Step 2 REQUIRED hard gate), validation checkpoints ("verify all selected prompt files exist", "Never start a batch until every selected prompt file exists on disk"), and feedback loops (retry failed items once, backup rule before overwrites) covering the batch generation risk. | 5 / 5 |
Progressive Disclosure | The body is a clear overview with 17 well-signaled, one-level-deep references, all verified to exist in the bundle, each with an explicit load condition (e.g., "Load references/codex-imagegen.md only for this branch", per-style specs read on demand). | 5 / 5 |
Total | 19 / 20 Passed |