Content
75%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 content is highly actionable, built around executable Membrane CLI commands and a well-sequenced connection workflow with explicit state-based validation. Its weaknesses are mild verbosity (product overview, a repeated command, vague filler) and the absence of validation guidance for destructive action/proxy calls.
Suggestions
Trim the opening paragraph explaining what ReferralRock is and remove the duplicated action-list command under 'Popular actions', which already appears in 'Searching for actions'.
Add a validation/checkpoint note before destructive operations (e.g., 'Confirm the action is non-destructive or dry-run proxy DELETE/PUT calls before committing') to lift workflow clarity.
Replace the filler line 'When to use which actions: Use action names and parameters as needed.' with concrete guidance on choosing between pre-built actions and the proxy.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Most of the body is efficient, executable CLI commands, but the intro paragraph explains what ReferralRock is (a concept Claude can infer), the 'Popular actions' section repeats the action-list command already shown, and filler like 'When to use which actions: Use action names and parameters as needed.' adds little, so it is mostly efficient with some tightening needed. | 3 / 5 |
Actionability | The body is dominated by copy-paste-ready commands (install, login, connection ensure/get, action list/run, request) plus a complete flag-options table, fully executable and covering the common cases. | 5 / 5 |
Workflow Clarity | The connection workflow is clearly sequenced (install, login, ensure, poll until ready) with explicit validation checkpoints (READY/BUILDING/CLIENT_ACTION_REQUIRED/CONFIGURATION_ERROR) and a feedback loop (re-poll after user action); the gap is that running or proxying potentially destructive actions has no validation step, keeping it just below a 5. | 4 / 5 |
Progressive Disclosure | No bundle files exist, so all content lives in one well-organized SKILL.md with clear section headers and one external docs link; structure is good and navigation is easy, but at ~130 lines of inline content it is longer than the simple-skill case that would score 5, leaving minor organization gaps. | 4 / 5 |
Total | 16 / 20 Passed |