Content
61%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.
The body is a well-organized, concrete instruction set for a single-purpose task: it specifies the exact file, entry syntax, section mapping, and output examples. Its weaknesses are redundancy between Instructions and Notes, the absence of an explicit verification step for the batch edit, and no executable git command for the blame-history discovery step.
Suggestions
Add a validation checkpoint after editing, e.g., 'Re-read the modified CODENOTIFY and confirm each new entry sits in the correct section, is alphabetically ordered, and duplicates no existing entry.'
Provide the concrete discovery command (e.g., git log --name-only --pretty=format: --author="<handle>" | sort -u) instead of just 'search through the repository's git blame history'.
Remove the Notes section's repetition of the Instructions and the trailing conversational line to tighten token usage.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient, but the Notes section restates instructions already given ('Maintain alphabetical or logical order', add entries to the appropriate section), the output examples duplicate the path-to-section mapping, and the closing conversational line is pure filler. | 3 / 5 |
Actionability | Concrete guidance throughout — exact target file, explicit entry formats ('path/to/file.ts @yourusername', 'dir/**'), a full path-to-section mapping, and worked output examples. Scored 4 rather than 5 because the git blame discovery step is described but not given as an executable command. | 4 / 5 |
Workflow Clarity | A clear numbered sequence exists (gather handle/aliases, search blame history, map paths to sections, add entries), but this batch file-edit operation has no validation checkpoint — nothing instructs verifying entries landed in the correct sections without duplicates after the edit, capping the score at 3. | 3 / 5 |
Progressive Disclosure | No bundle files exist and the single SKILL.md is cleanly sectioned (Instructions, What This Prompt Does, Expected Output Format, Notes) with easy navigation. At ~80 lines with some duplicated example content inline, structure is good but not fully optimally split, so 4 rather than 5. | 4 / 5 |
Total | 14 / 20 Passed |