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.
A strong, gate-driven process skill: every step carries concrete tool invocations, verbatim templates, and explicit validation with a feedback loop. The only real cost is token efficiency — the core gate and human-authority messages are stated two to three times over, which could be tightened without losing force.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Nearly all content is skill-specific craft rather than general knowledge, but the 'don't draft Design before framing' gate and 'acceptance is a human act' messages are each repeated across the intro, Step 1/Step 9, and Non-goals, and rhetorical framing like 'The failure this skill exists to prevent' could be trimmed. | 4 / 5 |
Actionability | Fully executable guidance throughout: concrete MCP calls (`write({ document: { path: "proposals/0007-async-export-pipeline.md", template: "proposal" } })`, `exec("ls -A proposals/")`, `audit({ path: ... })`) plus verbatim copy-paste templates for the framing block, alternatives structure, frontmatter, and recap. | 5 / 5 |
Workflow Clarity | A clearly sequenced 10-step (0–9) process with an explicit STOP gate requiring user confirmation before drafting, a Step 8 validation checklist ('audit returns clean — fix every finding'), and a defined error-recovery feedback loop — the top anchor's pattern exactly. | 5 / 5 |
Progressive Disclosure | Single self-contained file with no bundle files present; content is appropriately placed inline (a process skill's templates and gates belong in the main file), sections are clearly headed and navigable, and nothing needs external references. | 5 / 5 |
Total | 19 / 20 Passed |