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 genuinely executable Membrane CLI guidance with good connection-lifecycle handling, and is appropriately scoped as a single-file skill. Its weaknesses are moderate: a broken 'Step 2' cross-reference, no verification guidance around potentially destructive proxy requests, duplicated discovery commands, and placeholder sections (Popular actions, entity overview) that carry no content.
Suggestions
Fix the dangling reference: 'If the returned connection has state: READY, skip to Step 2' points to a step that doesn't exist — either number the workflow steps explicitly (1. Install, 2. Authenticate, 3. Connect, ...) or replace the reference with the actual next action ('skip to Searching for actions').
Add validation guidance for destructive proxy requests: before running POST/PUT/DELETE via membrane request, verify with a GET first, confirm IDs with the user, and check the response body for errors — the current proxy section sends arbitrary mutations with no checkpoints.
Either populate 'Popular actions' with the actual common Cafe 24 intents (e.g., list products, get orders by date) or remove it and the duplicated npx-form discovery command, keeping a single canonical form; also give the bare 'Cafe 24 Overview' entity list a purpose (e.g., which entities map to which action intents) or cut it.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient commands, but contains trimmable material: the introductory paragraph ("Think of it as a localized Shopify"), a near-duplicate action-discovery command (membrane vs npx @membranehq/cli@latest forms), an empty "Popular actions" section that adds no content, and padded prose like "This will burn less tokens". This matches 'mostly efficient but includes some unnecessary explanation or could be tightened' rather than anchor 4's minor-instances bar. | 3 / 5 |
Actionability | Concrete, executable commands appear throughout (npm install -g @membranehq/cli@latest, membrane connection ensure "https://www.cafe24.com/" --json, membrane action run <actionId> --connectionId=... --input '{...}') plus a proxy flags table — 'mostly executable guidance; concrete code or commands with minor gaps'. It falls short of 5 because there are no Cafe 24-specific worked examples (e.g., an actual intent query returning a real action for products or orders), and the 'Popular actions' section lists none. | 4 / 5 |
Workflow Clarity | The install → login → connection ensure → poll states → search actions → run sequence is present, and connection state handling is well covered (--wait polling, CLIENT_ACTION_REQUIRED, CONFIGURATION_ERROR → check error field). However, 'If the returned connection has state: READY, skip to Step 2' references a 'Step 2' that is never defined, and the proxy path permits arbitrary POST/PUT/DELETE requests with no validation or caution guidance, so the destructive/batch cap holds it at 'steps listed but validation gaps'. | 3 / 5 |
Progressive Disclosure | This is a single self-contained ~140-line file with no bundle files, organized under clear headers (Install → Authentication → Connecting → Searching → Running → Proxy → Best practices) appropriate to its size — 'good structure; most content is appropriately placed'. It misses 5 due to organization gaps such as the empty 'Popular actions' section and the bare nested entity list under 'Cafe 24 Overview' that serves no navigational purpose. | 4 / 5 |
Total | 14 / 20 Passed |