Content
85%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-crafted, appropriately sized body for a simple skill: the example table doubles as both format spec and template, tool selection is explicit with a fallback condition, and the classification criteria are domain-specific and concrete. The only improvements are trimming minor redundancy and specifying how to fetch reaction counts.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean: no concept explanations, and the table format is specified via a compact example rather than prose. Minor trimmable redundancy exists — the recommendations sentence re-lists 'fix now', 'queue', 'defer' already defined in the light criteria — so it falls just short of 'every token earns its place'. | 4 / 5 |
Actionability | Concrete, executable guidance is given: tool preference order (GitHub MCP, then 'gh issue list / gh issue view only when authenticated access is already configured'), a full example table, and explicit classification criteria per light. A minor gap remains: no guidance on how to retrieve reaction totals for the 'Reactions' column via the gh CLI, which is non-obvious. | 4 / 5 |
Workflow Clarity | This is a simple single-purpose skill (~25 lines) whose single action is unambiguous — inspect open issues, classify, render the table — with a tool-fallback checkpoint stated. It is read-only reporting with no destructive or mutating batch operation, so the validation cap does not apply and the simple-skill exception holds. | 5 / 5 |
Progressive Disclosure | The skill is under 50 lines, needs no external references (and references none — no references/, scripts/, or assets/ files exist in the bundle), and is organized into clear sections for scope, tool selection, table shape, classification, and output constraints. Per the simple-skill scoring note, this merits a 5. | 5 / 5 |
Total | 18 / 20 Passed |