Content
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured, highly actionable skill for PR review with clear workflow sequencing, explicit constraints, and concrete commands. Its main weakness is moderate verbosity from repeated prohibitions across multiple sections and a monolithic structure that could benefit from splitting reference material into separate files. The defensive guardrails (error handling, call budget, hard rules) are thorough and appropriate for an operation involving external API calls.
Suggestions
Consolidate the repeated prohibitions (e.g., 'do not manually call the PR Files API') into a single constraints section rather than restating them in hard rules, workflow steps, and call budget.
Consider extracting the detailed output format specification and review focus checklist into separate referenced files to reduce the main SKILL.md length.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is mostly efficient and avoids explaining concepts Claude already knows, but it's quite lengthy with repetitive prohibitions (e.g., 'do not manually call the PR Files API' appears in multiple forms across several sections). The call budget section partially duplicates constraints already stated in hard rules and workflow steps. | 2 / 3 |
Actionability | Provides fully concrete, executable commands with exact script paths, environment variable patterns, and specific invocation examples. The output format is precisely specified with tags, character limits, and section structure. The error block template is copy-paste ready. | 3 / 3 |
Workflow Clarity | The workflow is clearly sequenced (identify PR → collect data → targeted patches → raw files → optional checks → review output) with explicit validation checkpoints (PATCHES_SKIPPED triggers step 3, 404 handling, authentication failure handling) and a well-defined error recovery path. The call budget section acts as an explicit constraint checklist. | 3 / 3 |
Progressive Disclosure | The skill references external scripts (collect_pr.sh, fetch_targeted_patches.sh, fetch_raw_files.sh) which is good progressive disclosure, but no bundle files were provided to verify their existence. The SKILL.md itself is a long monolithic document that could benefit from splitting the output format specification or review focus checklist into separate reference files. | 2 / 3 |
Total | 10 / 12 Passed |