Content
82%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 highly actionable with complete, executable examples and a strong error-recovery workflow. It is mostly concise and well-structured, with only minor trimming and organization opportunities.
Suggestions
Mention scripts/common.js in the Files section (or note it is internal) so all bundle files are accounted for.
Trim the lengthy Gmail trash-counting example or move it to a references file, keeping a shorter representative exec example inline.
Add an explicit validation/checkpoint note to the batch listing examples (e.g. verify pagination completes and handle empty results) to push workflow clarity to 5.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly lean and assumes Claude's competence, but the long Gmail counting example and repeated 'node scripts/auth.js accounts' invocations across sections are minor padding that could be trimmed. | 4 / 5 |
Actionability | Fully executable, copy-paste-ready bash+JS examples cover the common cases, and the documented exec-context objects (auth, google, workspace.call/whoAmI) give concrete, specific guidance for arbitrary API calls. | 5 / 5 |
Workflow Clarity | The unauthorized/account-switch playbook gives a clear numbered sequence with an explicit feedback loop (enumerate → retry → re-login → retry original request); a couple of batch read examples lack explicit validation checkpoints, keeping it just below 5. | 4 / 5 |
Progressive Disclosure | The SKILL.md is a well-organized overview with a Files section that signals one-level-deep references to real bundle files (scripts/auth.js, scripts/workspace.js); the undocumented scripts/common.js and a couple of long inline examples are minor organization gaps. | 4 / 5 |
Total | 17 / 20 Passed |