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 concrete, mostly copy-paste-ready CLI guidance with a genuinely well-handled connection polling/state workflow, but it is undermined by dangling step references ('Step 2' is never defined, '1b' has no '1a'), duplicated discovery commands, and an unnecessary company-background preamble. Structure is good for a single-file skill, though a couple of sections could be split out.
Suggestions
Fix the workflow numbering: define an explicit "Step 2" (or renumber 'Connecting to Broadridge' as Step 1 and 'Searching for actions' as Step 2) so the "skip to Step 2" and "1b" references resolve.
Cut the opening Broadridge company-bio paragraph and the duplicated `action list` command in 'Popular actions' — keep one canonical discovery example.
Consider moving the proxy flag table and the clientAction state-machine explanation into a reference file, keeping SKILL.md focused on the get-connected-and-act flow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The opening blurb ("Broadridge is a financial technology company providing investor communications... serve banks, broker-dealers, asset managers") is background Claude does not need, and the `action list` command appears twice ('Searching for actions' and 'Popular actions'). Fits anchor 3 (mostly efficient, some unnecessary explanation or could be tightened); not 4 because the company-bio paragraph and duplicated discovery section are clear trim targets, and not 2 because the bulk of the body is lean, executable CLI content. | 3 / 5 |
Actionability | Commands are concrete and executable throughout (`npm install -g @membranehq/cli@latest`, `membrane connection ensure "https://www.broadridge.com/" --json`, `membrane action run <actionId> --input '{"key": "value"}'`) with a complete proxy flag table. Fits anchor 4 (mostly executable, minor gaps); not 5 because the instruction "skip to **Step 2**" references a step that is never defined, leaving a real execution-path gap, and placeholders like CONNECTION_ID are never tied to where the value comes from beyond one mention. | 4 / 5 |
Workflow Clarity | The connect → poll → act sequence exists and the connection state machine (READY / CLIENT_ACTION_REQUIRED with poll-again guidance) is a genuine checkpoint loop, but the workflow numbering is incoherent: "skip to **Step 2**" has no target, and "1b. Wait for the connection to be ready" sits under an unnumbered parent heading, so the intended step order must be inferred. Fits anchor 3 (sequence present but checkpoints/structure implicit); not 4 because the broken step references leave gaps in the defined sequence, and not 2 because a rough, mostly-followable sequence with state validation is present. | 3 / 5 |
Progressive Disclosure | With no bundle files, the single-file body is well-organized into clear sections (Install, Authentication, Connecting, Searching, Running actions, Proxy, Best practices) at reasonable length. Fits anchor 4 (good structure, most content appropriately placed, minor gaps); not 5 because content like the proxy flag table and the clientAction state-machine explanation are candidates for a reference file that would keep SKILL.md a leaner overview, and not 3 because nothing is buried and navigation is easy. | 4 / 5 |
Total | 14 / 20 Passed |