OpenPGx has 10 tools with basic naming and descriptions, but significant gaps in schema completeness and parameter documentation. Tool names follow verb_noun convention (upload_dna, load_profile, analyze_medication, get_gene_info, search_drugs, etc.), which is good. However, most tools lack detailed input parameter descriptions beyond the parameter name itself. For example, analyze_medication accepts 'drug_name' with description 'Brand or generic drug name (e.g., Zoloft, sertraline)', the example values in descriptions violate the rubric (LLMs reuse examples literally). Output schemas are not documented in the visible code. Parameter types are present (string, enum) but descriptions are minimal (10-30 chars), falling below the 72-char baseline. No error handling guidance visible. The upload_dna tool is the only one with a WRITE risk designation; others are READ_ONLY, which is appropriate. Overall, this is a fair-to-poor implementation typical of community servers (C-grade territory).

• Document output schemas for all tools. For get_gene_info, specify: {gene: string, activity_level: string, cpic_level: string, affected_drugs: string[], context: string}. For search_drugs, specify: {results: [{name: string, type: enum, category: string}], total: number, limit: number}.
• Remove example values from parameter descriptions. Replace 'Brand or generic drug name (e.g., Zoloft, sertraline)' with 'Brand or generic drug name. Call search_drugs() first if unsure of exact spelling.' Use enums for constrained values (search_drugs type: ['brand', 'generic', 'category', 'semantic']).
• Expand parameter descriptions to 50-150 chars. Current: 'Gene name' → Improved: 'Pharmacogene name (e.g., CYP2D6, CYP2C19, DPYD). Call get_gene_info() to list available genes and their clinical significance.'
• Add error handling guidance to tool descriptions. E.g., analyze_medication: 'If drug not found, returns empty result. Try search_drugs() to find the correct spelling or brand/generic variant.' get_gene_info: 'Returns null if gene not in PHARMACOGENE_CATALOG. Valid genes: CYP2D6, CYP2C19, CYP2C9, DPYD, TPMT, VKORC1, SLCO1B1, MTHFR, COMT, VDR, GLP1R, GIPR.'
• Add input validation and constraints. upload_dna: document max file size, supported formats (23andMe .txt, Genera .csv), and character encoding. search_drugs: add limit parameter (default 20, max 100) and document pagination. load_profile: validate code format (pgx-[a-z0-9]{6}) and return 'Profile not found' with recovery hint.

Inferred effective spec: <=2025-11-25.

• Relies on Stateful initialize / Mcp-Session-Id (removed; protocol is stateless) - make each request self-contained