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 solid, command-driven integration guide with strong actionability and decent structure. Its main issues are padding in the overview, a broken "Step 2" cross-reference with inconsistent CLI invocations ("npx @membranehq/cli" vs "membrane"), and missing validation around destructive operations.
Suggestions
Delete the padded overview (the "used by businesses and individuals" sentence, the Bot/Flow/Integration/User bullets, and "Use action names and parameters as needed.") and keep only the docs link.
Fix the workflow numbering: label the steps explicitly (Step 1: connect, Step 2: search/run actions) so "skip to Step 2" resolves, and use one consistent CLI invocation form.
Add a validation checkpoint before destructive actions (e.g. get-entity-item to confirm the target before delete-entity-item, or confirming with the user) to satisfy the destructive-operation feedback-loop requirement.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dominated by efficient commands and tables, but contains padding Claude does not need: "It's used by businesses and individuals looking to automate customer service, lead generation...", the content-free Overview bullets (Bot/Flow/Integration/User), and "Use action names and parameters as needed." This fits anchor 3 (mostly efficient with some unnecessary explanation) rather than 4, where only minor trimming would be needed. | 3 / 5 |
Actionability | Concrete, copy-paste-ready membrane commands cover install, login, connection, action search/run, and proxying, plus a 20-row action table and a proxy flag table. Not 5 because the run examples never reconcile the "<actionId>" placeholder with the table's "Key" column and no worked example passes realistic --input values. | 4 / 5 |
Workflow Clarity | A real sequence exists (install → login → connection ensure → poll for READY → search actions → run) with state-based feedback loops for CLIENT_ACTION_REQUIRED, but "skip to **Step 2**" references a heading that doesn't exist, and destructive actions (delete-entity, delete-entity-item) appear with no validation or confirmation steps. Per the destructive/batch cap, workflow clarity cannot exceed 3. | 3 / 5 |
Progressive Disclosure | As a single-file skill with no bundle, the body is well organized into install, auth, connection, actions, proxy, and best-practices sections with tables kept at reasonable inline length. It matches anchor 4 (good structure, content appropriately placed, minor organization gaps such as the mislabeled step headings); the under-50-line exception for a 5 does not apply to this ~160-line file. | 4 / 5 |
Total | 14 / 20 Passed |