Content
38%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill reads more like a product marketing page or README than an actionable skill for Claude. It covers breadth well — listing many capabilities and endpoints — but lacks depth: there are no executable code examples, no request/response formats, no error handling guidance, and no multi-step workflow instructions. The natural language 'examples' don't help Claude know how to actually call the API.
Suggestions
Add executable code examples (curl or Python) for at least the 3-4 most common operations (search, user lookup, tweet lookup, extraction), showing actual request construction with headers and expected response JSON structure.
Document multi-step workflows with validation checkpoints, especially for bulk extractions (estimate → start → poll status → retrieve results) and monitoring+webhooks (create monitor → register webhook → verify HMAC signature).
Remove or condense the 'When to Use This Skill' section — it largely duplicates the Capabilities table and adds ~15 lines of redundant content.
Create bundle reference files for detailed API documentation (request/response schemas, error codes, rate limits) and link to them from the main skill, rather than leaving the API reference as a shallow endpoint table.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill includes some unnecessary verbosity — the 'When to Use This Skill' section is a long list that largely restates what's in the Capabilities table. The overview paragraph also repeats the capabilities list. The examples section uses natural language prompts rather than executable code, which adds tokens without much value. However, the API reference table and setup instructions are reasonably tight. | 3 / 5 |
Actionability | The skill provides no executable code examples — only natural language prompt suggestions like 'Search X for tweets about claude code.' There are no actual API call examples (curl commands, Python requests, etc.), no request/response schemas, and no concrete code showing how to use the endpoints listed in the API reference table. The API table lists endpoints but doesn't show how to call them. | 2 / 5 |
Workflow Clarity | Setup steps are sequenced (install, get API key, set env var), but multi-step workflows like monitoring+webhooks or bulk extraction have no step-by-step guidance. There are no validation checkpoints — e.g., no guidance on checking if the API key works, verifying extraction completion, or handling rate limits/errors. The extraction workflow (estimate cost → start extraction → poll results) is implied by the endpoint table but never spelled out. | 3 / 5 |
Progressive Disclosure | No bundle files are provided, yet the skill has a lot of content inlined (full API reference table, capabilities table, extensive 'When to Use' list) that could benefit from separation. There are no references to detailed documentation files for advanced topics like webhook HMAC verification, extraction pagination, monitor configuration options, or error handling. Everything is surface-level in one file with no deeper resources. | 2 / 5 |
Total | 10 / 20 Passed |