Content
68%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 copy-paste Membrane CLI commands and a clear connection workflow, but it is padded by a redundant overview list and intro fluff, and it lacks verification steps for destructive proxy/action operations. Structure is good for a single-file skill.
Suggestions
Remove the redundant 'UpKeep Overview' bullet list (the four entities repeated in four orderings) and the filler line 'Use action names and parameters as needed.'
Add a verification checkpoint after running actions or destructive proxy requests (e.g., check the response `output`/error fields and confirm success before proceeding).
Trim the marketing-style intro sentence about industries and downtime; keep only what Claude needs to act.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The CLI sections are lean with copy-paste commands, but the intro fluff ('It's used by businesses across various industries to improve maintenance operations and reduce downtime'), the redundant 'UpKeep Overview' bullet list (the same four entities repeated in four orderings), and the filler line 'Use action names and parameters as needed.' are unnecessary. This is 'Mostly efficient but includes some unnecessary explanation' rather than score 4's tight, trimmed state. | 3 / 5 |
Actionability | It provides fully executable, copy-paste-ready commands covering the common cases — 'npm install -g @membranehq/cli@latest', 'membrane connection ensure "https://www.upkeep.com/" --json', 'membrane action run <actionId> --connectionId=CONNECTION_ID --json', and 'membrane request CONNECTION_ID /path/to/endpoint' — plus a flags table. Placeholders like CONNECTION_ID are standard and explained, so it clears score 5 rather than 4. | 5 / 5 |
Workflow Clarity | The connect-and-poll flow has explicit checkpoints and a feedback loop (poll until 'READY', re-poll after 'CLIENT_ACTION_REQUIRED'), but the proxy path exposes destructive methods (DELETE/PUT) and action runs have no verification step — 'The result is in the output field' with no confirm-success check. The rubric caps destructive/batch workflows lacking validation at 3, so it cannot reach score 4 despite the clear connection sequence. | 3 / 5 |
Progressive Disclosure | No bundle files exist, so all content is inline in one well-sectioned file with clear headers (Install, Authentication, Connecting, Searching, Running, Proxy, Best practices), matching 'Good structure; most content is appropriately placed.' It is not score 5 because the ~120-line file and the messy redundant overview are minor organization gaps, and not score 3 because navigation via headers is clear and references are not buried. | 4 / 5 |
Total | 15 / 20 Passed |