Two read-only tools with clear, domain-specific descriptions (194 - 250 chars, within baseline 34 - 392). Both have input schemas with typed parameters and descriptions. However, output schemas are completely undocumented, LLMs cannot plan downstream operations or extract structured data. Error handling is minimal (no recovery guidance, no categorization). Tool names are verb-noun but generic ('fcc_regulation', 'fcc_regulations_search'), no distinction between lookup-by-citation vs search-by-keyword in the names themselves. Parameter descriptions are adequate but lack format/constraint details (e.g., citation format examples are in the tool description, not the parameter schema). No pagination, no result limits documented, no idempotency guarantees.

• Document output schemas for both tools. Specify: for fcc_regulation, return {section: string, part: string, text: string, citation: string, effective_date: string}; for fcc_regulations_search, return {results: [{section, part, text_excerpt, citation}], total_count: number, limit: number, offset: number}.
• Add pagination to fcc_regulations_search: introduce 'limit' (default 20, max 100) and 'offset' (default 0) parameters. Document in tool description: 'Returns up to 20 matching sections per call. Use offset to paginate.'
• Enhance error messages with recovery guidance. E.g., 'eCFR API timeout (503). Retry in 5 seconds, or try a simpler citation format (e.g., "15.109" instead of "47 CFR 15.109(b)").'
• Formalize citation parameter constraints in schema: add 'pattern' field or enum examples. Or add a separate 'citation_format' enum parameter (lookup_by_section | lookup_by_part | lookup_by_name) to disambiguate intent.
• Rename tools to verb_noun convention: 'get_fcc_regulation' and 'search_fcc_regulations' for clarity and consistency with production baselines.