Content
78%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-structured skill body: executable quick-start examples, a validated multi-step workflow, and exemplary progressive disclosure through a conditionally-loaded reference table. The remaining gaps are minor — some repeated validation advice, a bullets list duplicating the frontmatter triggers, and error-recovery guidance that names the loop without showing it.
Suggestions
State the "validate with maxResults=1 before full execution" rule once (e.g., in the workflow or constraints) and remove the repetitions in the JQL sample comments and the MUST DO list.
Trim or replace the "When to Use This Skill" bullet list, which duplicates the frontmatter description's triggers, with content unique to the body (e.g., a decision pointer into the reference table).
Add a short error-recovery snippet (rate-limit backoff or network retry) or a direct pointer to the relevant reference section so the workflow's error-recovery step is concrete rather than named.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient — a compact workflow, tight example blocks, and scannable MUST DO / MUST NOT lists with no explanations of concepts Claude already knows. Not a 5 because of redundancy: the "validate with maxResults=1 first" advice is repeated three times (workflow step 3, the JQL sample comments, and the MUST DO list), and the "When to Use This Skill" bullets largely restate the frontmatter triggers. | 4 / 5 |
Actionability | Guidance is mostly executable: copy-paste-ready JQL and CQL samples, a complete MCP server JSON config with env-var placeholders, a security note on secrets, and concrete parameters ("50-100 items per page", "maxResults=1 probe"). Not a 5 because a few directives remain high-level ("Handle rate limits with exponential backoff", "Implement error recovery for network failures") without a snippet or pointer to a specific reference section for the common failure cases. | 4 / 5 |
Workflow Clarity | The 6-step Core Workflow is clearly sequenced with explicit validation checkpoints before risky operations — "validate with maxResults=1 before full execution" and "Confirm required scopes with a read-only probe before any write or bulk operation" — satisfying the batch/destructive validation requirement. Not a 5 because the error-recovery feedback loop is only named ("handle pagination, error recovery" in step 4) rather than shown as a validate-fix-retry cycle, and deploy's "test permissions" lacks a concrete check. | 4 / 5 |
Progressive Disclosure | The body is a genuine overview: quick-start examples stay inline, and all detailed guidance is split into five real one-level-deep reference files, each surfaced in a table with an explicit "Load When" condition. Cross-links between reference files are sibling navigation at the same level, not nested chains, matching the top anchor exactly. | 5 / 5 |
Total | 17 / 20 Passed |