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 delivers strong executable guidance — concrete CLI commands, a well-documented connection state machine, and a proxy fallback — organized in a logical single-file structure. Its weaknesses are padded intro prose, a workflow that references a nonexistent "Step 2" and offers no validation before destructive actions like Delete Collection, and an actions table with empty descriptions that leaves the most common operations undefined.
Suggestions
Fix the broken sequence: relabel the connection steps consistently (add a Step 1a/1b and an actual "Step 2: Search for actions" heading) so the "skip to Step 2" instruction resolves
Add validation guidance before destructive/batch actions — e.g. verify the collection ID with `get-collection-results` and confirm with the user before running `delete-collection`
Trim the marketing-style intro paragraph and the "When to use which actions: Use action names and parameters as needed" filler, and populate the empty Description column of the Popular actions table with what each action does
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is command-focused but includes unnecessary padding Claude does not need, e.g. "It helps teams organize tasks, track progress, and communicate effectively. It's typically used by project managers, team leads, and team members in various industries" and the throwaway "When to use which actions: Use action names and parameters as needed." — matching anchor 3 (mostly efficient but some unnecessary explanation). Not 4 because several sentences and the vague overview hierarchy add no actionable value. | 3 / 5 |
Actionability | Concrete executable CLI commands throughout (install, login, connection ensure, action list/run, proxy request with a flags table) with only minor gaps — the Popular actions table has an empty Description column and there is no fully-worked example of running a real action with actual parameters, matching anchor 4 rather than 5. | 4 / 5 |
Workflow Clarity | The connection flow has a good state machine with polling feedback (READY / CLIENT_ACTION_REQUIRED / CONFIGURATION_ERROR handling), but the skill exposes a destructive action ("Delete Collection") with no validation or verification guidance, which caps workflow clarity at 3 per the scoring notes. The sequence is also broken by the instruction "skip to **Step 2**" when no Step 2 heading exists (and section "1b" has no "1a"), keeping it at anchor 3 rather than 4. | 3 / 5 |
Progressive Disclosure | The skill is a single self-contained SKILL.md with no bundle files and no external references; sections are logically ordered (Install → Authentication → Connecting → Searching → Running → Proxy → Best practices) and the inline content (CLI flag table, connection states) is of appropriate size for a single file, matching anchor 4 (good structure, minor organization gaps such as the cryptic Document/Page/Template hierarchy list). | 4 / 5 |
Total | 14 / 20 Passed |