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 a solid, executable operational guide — its state-machine handling of connection readiness is a highlight. Its main defects are the dangling "Step 2" references that break the workflow's continuity and a few filler sections that dilute an otherwise efficient document.
Suggestions
Fix the broken step numbering: either add a labeled "Step 2" section (discover and run actions) or rewrite the two "skip to Step 2" references to point at the real section names, since currently no Step 2 exists.
Standardize on one CLI invocation form (`membrane` vs `npx @membranehq/cli`) or briefly explain when each applies, and drop the content-free 'Bearer Overview' entity list and "Use action names and parameters as needed." line.
Merge the 'Popular actions' section into 'Searching for actions' — it currently repeats the same discovery command without adding anything.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly lean and command-driven (install, login, connection ensure, the proxy flag table), with only minor trimmable filler: the bare entity bullet list under 'Bearer Overview', "Use action names and parameters as needed.", and the Agent Types aside. Not 5 because those tokens don't earn their place; not 3 because there is no explanation of concepts Claude already knows. | 4 / 5 |
Actionability | Concrete, copy-paste-ready commands cover the common cases (membrane login, connection ensure/get with --wait, action list --intent, action run --input, request with flags). Not 5 because of minor gaps: `membrane` and `npx @membranehq/cli` invocations are mixed without explanation, and the `connection ensure` output fields are referenced but never shown. | 4 / 5 |
Workflow Clarity | A rough sequence exists (install → login → connect → poll state → discover → run → proxy) with genuinely good state-branching for CLIENT_ACTION_REQUIRED, but "skip to **Step 2**" appears twice while no Step 2 section exists — the numbering jumps from an unlabeled step to "#### 1b", leaving the intended flow incoherent at exactly the point it matters. Not 4 because of this dangling reference; not 2 because the sequence and state validation checkpoints are present. | 3 / 5 |
Progressive Disclosure | This is a single-file skill (no references/, scripts/, or assets/ directories exist), and the body is well-sectioned with clear headers, keeping content appropriately inline. Not 5 because at ~145 lines it exceeds the simple-skill threshold and the 'Popular actions' section merely duplicates the 'Searching for actions' command with no added value. | 4 / 5 |
Total | 15 / 20 Passed |