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, largely executable CLI-integration guide with good error-state handling and a clear overall workflow. Its main weaknesses are avoidable redundancy (duplicated discovery command, content-free overview list) and the lack of any concrete filled-in AppTweak example, plus a dangling 'Step 2' reference from unlabeled step numbering.
Suggestions
Remove the duplicated 'action list' command (keep it in one place) and either give the 'AppTweak Overview' list meaning or delete it.
Fix the step numbering: label 'Step 1'/'Step 2' explicitly or replace 'skip to **Step 2**' and the '1b' heading with descriptive section references.
Add one end-to-end concrete example — e.g. a filled '--intent "get keyword rankings for an app"' query, a real action id, and its run command — and consider moving the clientAction schema detail and proxy flags table to a reference file to slim the main body.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly tight operational guidance, but the 'AppTweak Overview' bullet list (App / Keyword / Competitor / Review / Report) carries no information, the 'action list' discovery command is stated twice ('Searching for actions' and 'Popular actions'), and the headless-login flow is explained twice (browser vs. URL). Anchor 3 ('Mostly efficient but includes some unnecessary explanation or could be tightened'); not 4 because the duplication and content-free overview list are avoidable padding. | 3 / 5 |
Actionability | Commands are concrete and executable throughout: 'membrane connection ensure "https://apptweak.com" --json', 'membrane action run <actionId> --connectionId=... --input ...', and a full proxy flags table. Not 5 because no filled-in AppTweak example exists — no sample --intent query, real action id, or real endpoint path — so the common cases are covered only by placeholder-form commands; anchor 4 ('Mostly executable guidance; concrete code or commands with minor gaps'). | 4 / 5 |
Workflow Clarity | The sequence install → login → connection ensure → poll until READY → search actions → run/proxy is clear, and the connection phase has strong validation (state polling with --wait, CLIENT_ACTION_REQUIRED handling with clientAction fields, CONFIGURATION_ERROR/SETUP_FAILED branches). Not 5: the text says 'skip to **Step 2**' and uses a '1b' heading, but no Step 1/2 is ever labeled, and there is no post-run output verification; anchor 4 ('Clear sequence with most checkpoints present; minor validation gaps'). | 4 / 5 |
Progressive Disclosure | No bundle files exist (no references/, scripts/, or assets/), so everything is inline in one well-sectioned file with a logical header hierarchy. The inline clientAction schema explanation and proxy flags table could plausibly live in a reference file, but placement is reasonable for a single-purpose CLI skill — anchor 4 ('Good structure; most content is appropriately placed; minor organization gaps'); not 5 given the length exceeds the simple-skill threshold and some detail could be split out. | 4 / 5 |
Total | 15 / 20 Passed |