Content
76%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.
This is a highly actionable integration skill with copy-paste-ready CLI commands, a concrete action catalog, and a well-handled connection state machine with polling and error states. Its weaknesses are the missing validation/confirmation guidance for destructive and batch actions (which caps workflow clarity) and a few padded or marketing-flavored sentences that could be trimmed.
Suggestions
Add a validation checkpoint before destructive or batch actions — e.g. "Before running delete-contact or send-campaign, confirm the target with the user and re-fetch the record with get-contact to verify it's the right one".
Fix the broken step numbering: "skip to Step 2" references a step that is never defined — label the setup steps explicitly (Step 1: connect, Step 2: search/run actions).
Trim filler such as "It helps users network and exchange contact information more efficiently" and the URL-normalization aside, and correct "This will burn less tokens" to "use fewer tokens".
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dominated by efficient, copy-paste commands and tables, but contains minor over-explanation that could be trimmed: "It helps users network and exchange contact information more efficiently", the URL-normalization aside, and awkward phrasing like "This will burn less tokens". It is more efficient than the score-3 anchor's unnecessary commentary but not the lean every-token-earns-its-place level of 5. | 4 / 5 |
Actionability | Every workflow step ships an executable command: "npm install -g @membranehq/cli@latest", "membrane login complete <code>", "membrane connection ensure \"https://amcards.com/\" --json", "membrane action list --intent \"QUERY\" --limit 10 --json", "--input '{\"key\": \"value\"}'", plus a 17-action catalog and a proxy flag table. Commands are copy-paste ready and cover the common cases. | 5 / 5 |
Workflow Clarity | The setup sequence is clear with a well-specified connection state machine (READY / BUILDING / CLIENT_ACTION_REQUIRED / CONFIGURATION_ERROR) and a polling loop, but the skill exposes destructive and batch operations ("Delete Contact", "Send Campaign", mailings) with no pre-action confirmation or validation guidance, which caps this dimension at 3 per the rubric. The dangling "skip to Step 2" reference to a step that is never labeled adds a further gap. | 3 / 5 |
Progressive Disclosure | No bundle files exist; the skill is a single well-sectioned SKILL.md (Overview, Install, Authentication, Connecting, Searching, Actions, Proxy, Best practices). The under-50-line exception for a 5 does not apply at ~155 lines, and content like the 17-row actions table and dangling step references keep it at "good structure with minor organization gaps" rather than fully clear one-level-deep navigation. | 4 / 5 |
Total | 16 / 20 Passed |