Content
57%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 mostly executable, copy-paste-ready CLI guidance with a genuinely useful connection-state decision tree. Its main weaknesses are padding (marketing intro, empty entity list), broken internal step references ('Step 2', '1b' with no numbered steps), and inlining of generic Membrane CLI reference material that belongs in a separate file.
Suggestions
Fix the dangling cross-references: either number the workflow steps (Step 1: install/login, Step 2: search actions, ...) so 'skip to Step 2' and '1b' resolve, or replace them with section names like 'Searching for actions'.
Move the proxy flags table and clientAction state documentation into a reference file (e.g., references/membrane-cli.md) and keep only the core workflow in SKILL.md; this would also cut the marketing intro and the content-free entity bullet list.
Add one or two concrete end-to-end examples with real values, e.g., `membrane action list --intent "list contracts" --connectionId=... --json` followed by a sample `action run` invocation, so placeholders like QUERY and CONNECTION_ID are grounded.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The command blocks are tight, but there is removable padding: "It's used by businesses of all sizes to create, negotiate, and manage contracts more efficiently" (marketing fluff), an entity-name bullet list (Contract / Counterparty / Field / Template / Workspace / User) with no explanatory content, and "so you can focus on the integration logic rather than auth plumbing". This matches 'Mostly efficient but includes some unnecessary explanation or could be tightened'. | 3 / 5 |
Actionability | Concrete, executable commands throughout — `npm install -g @membranehq/cli@latest`, `membrane connection ensure "https://commonpaper.com" --json`, `action run` with `--input` JSON, and a proxy flags table — with minor gaps: no example action IDs or Common Paper endpoints for the proxy path. Matches 'Mostly executable guidance; concrete code or commands with minor gaps'; not a 5 because the placeholder-only examples (QUERY, CONNECTION_ID) never show a real Common Paper operation. | 4 / 5 |
Workflow Clarity | A sequence is present (install → login → connection ensure → poll states → search actions → run → proxy) with a good feedback loop for CLIENT_ACTION_REQUIRED, but "skip to **Step 2**" and "#### 1b" reference a numbered step structure that does not exist in the document, and there are no verification checkpoints after running actions. Matches 'Steps listed but validation gaps; sequence present but checkpoints missing or implicit'. | 3 / 5 |
Progressive Disclosure | No bundle files exist and the ~141-line SKILL.md has reasonable section headers, but generic Membrane CLI mechanics (login, connection lifecycle, proxy flags) that are common to all Membrane app skills are inlined rather than split into a shared reference file, matching 'content that should be separate is inline'. Not a 4 because the proxy flag table and clientAction state documentation are reference material inlined into the main file. | 3 / 5 |
Total | 13 / 20 Passed |