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 for the Membrane CLI workflow — install, auth (including headless), connection setup with a well-handled state machine, action discovery/run, and a proxy fallback. Its main weaknesses are a broken "Step 2" cross-reference, no validation of action-run results, a padded entity list that adds ~45 lines of unclear value, and a "Popular actions" section that lists no actions.
Suggestions
Fix the dangling references: "skip to Step 2" (lines 110, 124) points at a nonexistent section — introduce explicit numbered step headings (1. Connect, 2. Discover actions, 3. Run actions) and target them consistently.
Add error handling and result validation for the run step, e.g. checking the `output`/`error` fields of `membrane action run` responses and how to recover, mirroring the good state-branching already present for connections.
Trim the 43-item entity list (or move it to a reference file with one-line descriptions of which actions each entity exposes), and either populate "Popular actions" with real examples (e.g. task.create, driver.list) or remove the section.
Reconcile the invocation style: line 117 uses "npx @membranehq/cli" while everywhere else uses "membrane".
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient — the CLI install/auth/connection/proxy sections are lean commands — but the marketing-style intro ("helps businesses orchestrate and optimize their last-mile delivery operations... in-house fleets, 3PLs") and the 43-item flat entity list capped by the filler line "Use action names and parameters as needed" are unnecessary padding that could be tightened. It is not level 2 because there is no extensive explanation of concepts Claude already knows; the majority of tokens are executable guidance. | 3 / 5 |
Actionability | Concrete copy-paste commands throughout ("npm install -g @membranehq/cli@latest", "membrane connection ensure \"https://www.bringg.com/\" --json", "membrane action run <actionId> --connectionId=CONNECTION_ID --input '{\"key\": \"value\"}' --json") plus a clear proxy flag table. Minor gaps keep it from 5: no worked example with a real Bringg action and values, the "Popular actions" section contains no actual actions, and line 117 switches to "npx @membranehq/cli" while the rest uses "membrane". | 4 / 5 |
Workflow Clarity | The connection flow is well checkpointed (poll with --wait, explicit READY / CLIENT_ACTION_REQUIRED / CONFIGURATION_ERROR branching, re-poll after user action), but "skip to **Step 2**" appears twice (lines 110 and 124) pointing at a section that does not exist in the document, and there is no validation or error-recovery guidance for the action-run step itself. The sequence is present with real gaps, matching 'steps listed but validation gaps' rather than the minor-gaps level 4. | 3 / 5 |
Progressive Disclosure | A single self-contained file with well-ordered, clearly headed sections and no bundle files (references/, scripts/, assets/ absent) — none strictly needed since discovery is via the CLI. Structure is good and navigation easy, but at ~180 lines it exceeds the simple-skill exception, and the entity list and proxy flag table could live in a reference file, so it falls short of the well-split level 5. | 4 / 5 |
Total | 14 / 20 Passed |