Content
81%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-sequenced procedural skill with strong validation gates around an external create operation. Its main weaknesses are a padded prerequisites paragraph and a monolithic structure with no reference-file offloading.
Suggestions
Tighten the Prerequisites MCP paragraph to a sentence or two; drop the enumeration of MCP failure modes that Claude can infer.
Move the component-label mapping table and the Zeebe label-split guidance into a references/ file (e.g. COMPONENT_LABELS.md) and link to it from Step 3 to reduce inline bulk.
Consider extracting the GraphQL issue-type and sub-issue registration snippets in Step 7 into a scripts/ helper so the core flow stays scannable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient tables and commands, but the Prerequisites MCP paragraph over-explains credential and failure semantics ('not immune to failures of their own (their own rate limits, service-side restrictions)') that could be trimmed. | 3 / 5 |
Actionability | Provides copy-paste-ready gh and gh api graphql commands, a mktemp body-file pattern, and explicit template/label mapping tables covering the common cases. | 5 / 5 |
Workflow Clarity | Steps 1–7 are clearly sequenced with explicit validation checkpoints: duplicate search (1.5), an approval gate before creation ('Do NOT call gh issue create yet'), redaction, and a change-revision feedback loop. | 5 / 5 |
Progressive Disclosure | Well-sectioned single file with one clearly signaled one-level reference (AGENTS.md) and the grill-me skill, but at ~240 lines it inlines mapping tables that could be split into reference files and ships no bundle. | 4 / 5 |
Total | 17 / 20 Passed |