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 body is a well-structured, actionable CLI integration guide with concrete commands and explicit connection-state feedback loops. Its main weakness is conciseness: it opens with an unnecessary explanation of what Product Hunt is and includes minor fluff that Claude does not need.
Suggestions
Remove the opening paragraph explaining what Product Hunt is ("Product Hunt is a website where users discover and share new products...") — Claude already knows this.
Delete low-value filler such as "so you can focus on the integration logic rather than auth plumbing" to tighten token efficiency.
Trim the redundant "Product Hunt Overview" nested singular/plural list, which adds no actionable information beyond the action-search guidance.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient with concrete commands, but includes unnecessary explanation Claude already knows, e.g. "Product Hunt is a website where users discover and share new products. It's primarily used by tech enthusiasts, entrepreneurs, and early adopters..." and the fluff "so you can focus on the integration logic rather than auth plumbing." It is not a 4 because the what-is-Product-Hunt paragraph is clear over-explanation, and not a 2 because the bulk is lean executable guidance rather than heavily padded. | 3 / 5 |
Actionability | Provides copy-paste-ready commands covering the common cases — install (npm install -g @membranehq/cli@latest), login, connection ensure, action list/run with --input JSON, and request proxying with a flag table. It is not a 4 because the examples comprehensively cover install through execution with concrete flags and JSON output, matching the fully-executable anchor. | 5 / 5 |
Workflow Clarity | There is a clear sequence (install → authenticate → ensure connection → wait for ready → search actions → run/proxy) with feedback loops ("poll again", "Keep polling until state is no longer BUILDING") and explicit state handling (READY/BUILDING/CLIENT_ACTION_REQUIRED/CONFIGURATION_ERROR). It is not a 5 because the "Step 2" referenced in the text is not visibly labeled as a step, creating a minor sequencing gap; not a 3 because validation checkpoints and error-recovery loops are present. | 4 / 5 |
Progressive Disclosure | No bundle files exist, and the single SKILL.md is organized into well-labeled sections (Overview, Working with Product Hunt, Install, Authentication, Connecting, Searching, Popular actions, Proxy, Best practices) with no nested references. It is not a 5 because the skill exceeds 50 lines and the "Product Hunt Overview" nested list (Posts/Post, Comments/Comment) is low-value padding that could be trimmed; not a 3 because structure is clear and content is appropriately placed. | 4 / 5 |
Total | 16 / 20 Passed |