Content
50%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 skill is highly actionable with concrete, executable Membrane CLI commands and a solid connection workflow, but its quality is dragged down by a large redundant keyword dump and missing validation for destructive operations.
Suggestions
Delete the ~150-word bullet keyword dump under Amplication Overview; it is pure padding that adds no guidance.
Add a validation/verification step after destructive `action run` or `membrane request -X DELETE/PUT` calls (e.g. re-query the resource to confirm the change) so the workflow clears the destructive-operation cap.
Trim explanatory prose that states what Claude already knows (e.g. what Amplication is, that Membrane handles auth) to keep the body lean.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body opens with roughly 150 single-word bullet verbs (many duplicated, e.g. Monitor, Analyze, Optimize, Scale, Backup, Restore appearing twice) that explain nothing Claude does not already know and add severe padding. | 1 / 5 |
Actionability | It provides copy-paste-ready CLI commands (install, login, connection ensure, action list/run, request) with concrete flags, JSON examples, and a full options table covering the common cases. | 5 / 5 |
Workflow Clarity | The connection setup has a clear poll-until-ready sequence with state checks and a feedback loop, but destructive operations like `action run` and `membrane request` (DELETE/PUT) lack validation checkpoints, capping the score. | 3 / 5 |
Progressive Disclosure | There are no bundle files and the SKILL.md uses reasonable section headers, but the inline keyword dump is content that belongs nowhere and signals poor organization rather than a clean overview. | 3 / 5 |
Total | 12 / 20 Passed |