Content
86%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 highly actionable, well-structured body built on executable gh CLI commands with good error-recovery branching and clear guardrails. Its only notable slack is mild redundancy between the frontmatter description and the "When to Use" section, plus a few unhandled edge cases in the workflow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is efficient — executable gh commands with jq filters and no explanations of concepts Claude already knows — but the "When to Use" section largely restates the frontmatter triggers and the sample table block adds padding that could be trimmed, keeping it below the lean-every-token-earns-its-place anchor. | 4 / 5 |
Actionability | Every step ships copy-paste-ready commands (gh pr view/checks, gh run view --log-failed, gh run rerun --failed) with specific --json/--jq flags, plus ready-made user-facing response scripts and a concrete output format, fully covering the common cases. | 5 / 5 |
Workflow Clarity | The five-step sequence is clear with strong fallback branches (no PR given → find it; no PR exists → list branch runs) and escalation/retry paths for failures; minor validation gaps (e.g., no handling of gh errors, auth failures, or empty check output) keep it below the explicit-checkpoint anchor, and the safety checkpoint "Re-run jobs without asking" guards the one semi-destructive action. | 4 / 5 |
Progressive Disclosure | The skill is self-contained with no bundle files, yet content is appropriately sized — nothing that belongs in a separate file is inlined — and sections (When to Use, Process, Re-running Jobs, Integration, Do NOT) are well-organized with clear cross-skill navigation links. | 5 / 5 |
Total | 18 / 20 Passed |