Content
71%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 clean, well-structured instruction skill for a remote MCP tool with concrete configuration and a usable example. Main weakness is redundancy across sections and a minor validation/parameter gap.
Suggestions
Deduplicate the returned-fields list — state the 25+ fields once (in Usage) rather than repeating methods/results/sample sizes/quality across the description, Overview, and Usage sections.
Add a brief note on verifying the BGPT MCP server connected successfully before issuing the first search_papers call, since it is a remote server that can fail to connect.
Document the search_papers tool's key parameters (e.g., query, filters, max results) so the agent can construct the call directly rather than from a natural-language example alone.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient but repeats the returned-fields list across the description, Overview, and Usage sections, and lightly over-explains what a literature database is ('Unlike traditional literature databases that return titles and abstracts'); could be tightened. | 3 / 5 |
Actionability | Provides copy-paste-ready MCP config JSON and an executable npm command plus a concrete example query, but does not surface the search_papers tool's parameters/schema, leaving a minor gap. | 4 / 5 |
Workflow Clarity | Clear Setup-then-Usage sequence for a simple single-tool skill, but there is a minor validation gap: no checkpoint confirming the remote MCP server actually connected before calling search_papers. | 4 / 5 |
Progressive Disclosure | Well-organized into clearly signaled sections (Overview, When to Use, Setup, Usage, Pricing) with no nested references; a self-contained remote-tool skill appropriately kept in one file. | 5 / 5 |
Total | 16 / 20 Passed |