Content
67%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.
A solid, executable CLI integration guide: real commands throughout, explicit connection-state handling with feedback loops, and sensible single-file organization. Its weaknesses are token waste in the empty overview and duplicated actions sections, and a dangling "Step 2" reference that breaks the workflow's navigation.
Suggestions
Delete the "Contact Enhance Overview" hierarchy and the "Use action names and parameters as needed." line — they carry no information.
Merge "Popular actions" into "Searching for actions" and either list a few real Contact Enhance actions with example inputs or drop the section, since it currently repeats the same command verbatim.
Replace the two dangling "skip to **Step 2**" references with a named target (e.g. "skip to **Searching for actions**") or add explicit step numbering across the workflow sections.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly lean, command-first prose, but three spots waste tokens: the "Contact Enhance Overview" hierarchy that lists only "Contact / Enhancement Request", the filler line "Use action names and parameters as needed.", and the "Popular actions" section repeating the identical command already shown under "Searching for actions". That matches anchor 3 (mostly efficient but could be tightened) rather than 4 given the multiple unnecessary sections. | 3 / 5 |
Actionability | Nearly every instruction is a copy-paste-ready command — `npm install -g @membranehq/cli@latest`, `membrane connection ensure "https://contactenhance.com/" --json`, `membrane action run <actionId> --input '{...}' --json`, plus a complete proxy flag table. It falls short of 5 only because the "Popular actions" section lists no actual actions and the inputSchema/outputSchema workflow is never demonstrated with a concrete example. | 4 / 5 |
Workflow Clarity | The install → authenticate → connect → discover → run sequence is clear and the connection state machine provides explicit checkpoints: poll with `--wait`, handle `clientAction` per type, re-poll after user action, and check the `error` field on failure. The gap keeping it at 4 rather than 5 is that both "skip to **Step 2**" references are dangling — no Step 2 exists in the document. | 4 / 5 |
Progressive Disclosure | The skill is a single appropriately sized file with clear section headers, no bundle files, and no references to nonexistent paths, so all content is correctly placed inline. Minor organization gaps (the pointless overview hierarchy and the duplicated actions section) keep it at anchor 4 rather than 5. | 4 / 5 |
Total | 15 / 20 Passed |