IBGE Data Server provides 6 read-only tools for querying Brazilian statistical data. All tools have descriptions and input schemas are visible in the code. However, several structural issues limit quality: (1) Tool naming uses Portuguese verb forms (listar_, obter_, consultar_, buscar_) which diverges from the English verb_noun convention that LLMs expect for parsing intent. (2) Parameter descriptions in the schema include example values (e.g., 'ex: 1705, 1712') which LLMs tend to reuse literally rather than adapt to context, this violates the pattern:constrained-input guidance. (3) Output schemas are not documented anywhere in the visible code or descriptions; LLMs cannot infer what fields to expect from tool responses. (4) The server has no error handling patterns visible, no recovery guidance, no categorization of errors as retryable vs. fatal. (5) Parameters like 'variavel', 'periodos', 'classificacao', and 'localidades' accept complex string formats ('214|1982', '-6', '226[4844]|218[4780]') without clear enum constraints or regex patterns, this invites hallucinated input from LLMs. (6) The 'view' parameter on consultar_dados_variaveis is loosely documented; 'OLAP', 'flat', 'default' should be declared as an enum. Tools are read-only (good) but the composition pattern shows that listar_agregados and buscar_agregados_por_termo appear to do similar work, LLMs will conflate them. Parameter descriptions are present but average 50-80 chars, below the 72-char baseline for A+ tools. Overall, the server is functional but lacks LLM-optimized documentation and output schema visibility.

• Rename tools to English verb_noun convention: list_aggregates, get_aggregate_metadata, get_localities, get_periods, query_variable_data, search_aggregates_by_term. This aligns with LLM expectations and the 549-tool baseline.
• Remove example values from parameter descriptions and replace with formal constraints. For agregado_id, use 'Integer >= 1' or document the valid range from IBGE API. For periodos, specify 'Format: -N (last N periods), or YYYYMM-YYYYMM (date range)' with regex pattern if possible.
• Document output schemas for every tool. Add a section to each tool description: 'Returns: Array of objects with fields: id (integer), name (string), metadata (object). See IBGE API docs for full field details.' This enables LLMs to extract and chain data.
• Declare parameter enums formally. For 'view' on consultar_dados_variaveis, use enum: ['OLAP', 'flat', 'default']. For 'nivel' on obter_localidades, enumerate valid geographic levels (N1, N2, N6, N7) with descriptions.
• Add error handling descriptions to all tools. E.g., 'If agregado_id is invalid, the API returns 404. This is a user error, ask for a valid aggregate ID using list_aggregates or search_aggregates_by_term first. Do not retry.'
• Consolidate search functionality. Either merge listar_agregados and buscar_agregados_por_termo into a single search_aggregates tool with optional term/filter parameters, or clearly document when each should be used. Reduce LLM ambiguity.
• Add pagination and result limits to tools returning lists. E.g., 'listar_agregados returns max 20 results; use offset and limit parameters for pagination.' This prevents context window exhaustion.

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

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry