Content
82%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.
This is a dense, high-signal instruction skill: exact gh commands with a cloud-session fallback, explicit decision criteria, built-in verification steps, and clear routing to sibling skills. The only notable gaps are the headerless wall-of-prose organization and a final paragraph that drifts into abstract maintainer-session jargon (EGM, Cloud node) without concrete steps.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean throughout: no explanation of what gh, FastMCP, or triage concepts are; every sentence carries a directive ("Do not rank solely by recency, comment count, or how small a proposed patch looks", "check for a revert first"). It assumes Claude's competence and every token earns its place, matching anchor 5. | 5 / 5 |
Actionability | Concrete, copy-paste-ready commands dominate: exact gh invocations with --json field lists, a REST fallback URL ("gh api \"repos/PrefectHQ/fastmcp/pulls?state=all&per_page=100&page=N\""), and specific decision rules (want/marginal/don't want, cleanup-pass criteria). The final paragraph's batch/maintainer guidance ("Use upstream EGM in the maintainer session") is comparatively abstract and unexecutable, keeping it below anchor 5. | 4 / 5 |
Workflow Clarity | A clear sequence runs from reading AGENTS.md and the contribution policy, through a bounded inventory and PR linking, to verification checkpoints (trace code and git history, distinguish fixed/released/proposed, check for reverts) and a defined output. It is not 5 because there are no explicit feedback loops or error-recovery steps for the verification phase (e.g., what to do when evidence contradicts a pick), though the skill is read-only so the destructive-operation cap does not apply. | 4 / 5 |
Progressive Disclosure | At ~30 lines it stays within SKILL.md appropriately, with clearly signaled one-level references to sibling skills ([fix-issue](../fix-issue/SKILL.md), [review-issue](../review-issue/SKILL.md)) and the contributing guide; no nested reference chains. It falls short of anchor 5 because there are no section headers at all — the guidance is dense flowing prose a reader must parse linearly — and some reference paths cannot be resolved from the skill's own directory. | 4 / 5 |
Total | 17 / 20 Passed |