Content
68%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 is an efficient, largely executable CLI integration guide with good section structure and concrete state-handling for the connection flow. Its main weaknesses are an undefined "Step 2" cross-reference, missing validation around the destructive delete-campaign action, and the absence of a complete worked example.
Suggestions
Label the workflow steps explicitly (Step 1: connect, Step 2: find/run actions) or remove the "skip to **Step 2**" references that currently point to a step that is never defined.
Add a validation checkpoint before destructive operations — e.g. verify the campaign id via get-campaign-status and confirm with the user before running delete-campaign.
Include one end-to-end worked example showing the connection id being extracted from the `connection ensure` output and substituted into an `action list`/`action run` invocation.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Command-first sections are efficient and assume competence ("Add `--json` to any command", "The result is in the `output` field"), with only minor trimmable filler such as "Use action names and parameters as needed", the marketing-oriented intro paragraph, and a truncated table cell ("...l..."). | 4 / 5 |
Actionability | Nearly all guidance is copy-pasteable (`npm install -g @membranehq/cli@latest`, `membrane connection ensure "https://campaigncleaner.com/" --json`, `membrane action run <actionId> --input '{"key": "value"}'`, proxy flag table), but there is no end-to-end worked example showing how the connection id is captured from output and substituted into subsequent commands. | 4 / 5 |
Workflow Clarity | A rough sequence exists (install → login → connection ensure → poll for READY → search actions → run) with state-based checkpoints (`CLIENT_ACTION_REQUIRED`, `CONFIGURATION_ERROR`), but "skip to **Step 2**" references a step that is never defined, and the destructive `delete-campaign` action is listed with no validation/confirmation guidance — the destructive-operation cap applies. | 3 / 5 |
Progressive Disclosure | A single well-sectioned SKILL.md (~150 lines) with clear headers and a small actions table; no nested references and no dead links, though the popular-actions table and proxy details could live in a one-level-deep reference file if the skill grows. | 4 / 5 |
Total | 15 / 20 Passed |