What problem does it solve? ToolUniverse research skills often produce noisy, incomplete, or process-heavy reports due to silent tool failures, wrong API parameters, missing disambiguation, and ungraded evidence. This Skill provides a structured methodology and checklist to diagnose and fix those quality problems. ## Core Features & Use Cases - Tool Contract Verification: Verify tool parameters via get_tool_info(), maintain corrections tables, and handle versioned versus unversioned identifiers (e.g., GTEx Ensembl IDs). - Research Workflow Patterns: Apply foundation-data-first queries, disambiguation phases with collision filtering, citation expansion, and retry/fallback chains for flaky APIs like PubMed and EuropePMC. - Report Quality Standards: Enforce T1-T4 evidence grading, quantified completeness minimums, mandatory sections, aggregated Data Gaps sections, and narrative reports separated from JSON bibliographies. - Use Case: When a gene-target research skill returns empty GTEx expression data and a report full of search logs, use this Skill to identify the versioned-ID issue, add fallback chains, and restructure the output into a content-focused report. ## Quick Start Review my ToolUniverse research skill and optimize it using the evidence grading, disambiguation, and completeness checklist patterns.
