Skip to main content
Glama

MCP Compras.gov.br

by io.github.opedrosoares

compras_buscar_contratacoes_similares

Read-onlyIdempotent

Federa Dados Abertos + PNCP buscando contratações similares.

Composição: consulta os itens de contratações 14.133 no Dados Abertos (/modulo-contratacoes/2_, filtrando por codItemCatalogo e só itens com resultado) + publicações PNCP do período, deduplica pelo número de controle PNCP e devolve os max_resultados mais recentes. Insumo para mapear benchmarks de outros órgãos.

O recorte por CATMAT/CATSER vale para a perna Dados Abertos. A perna PNCP é best-effort por modalidade e não aceita filtro por item de catálogo — por isso amostra_dados_abertos e amostra_pncp vêm separadas no payload.

Atenção latência: chama o PNCP em 3 modalidades (Pregão, Dispensa, Concorrência) em paralelo. Cada chamada PNCP costuma levar 30-60s — o tempo total da composta tende a 60-90s quando o cache está frio. Com Redis configurado as chamadas seguintes voltam em <1s.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
ufNoFiltro opcional por UF.
codigo_catmatNoCódigo CATMAT do item. Mutuamente exclusivo com codigo_catser.
codigo_catserNoCódigo CATSER do serviço. Mutuamente exclusivo com codigo_catmat.
periodo_mesesNoJanela de busca em meses contados de hoje para trás.
max_resultadosNoMáximo de contratações similares a retornar (deduplicadas).

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault

No arguments

Schema Changelog

Changes observed during successful MCP inspections. Dates show when Glama detected each change.

  1. First observed

TDQS

A4.7/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond the readOnly/idempotent/destructive annotations, the description discloses substantial behavioral detail: the composite data-source flow, deduplication by PNCP control number, the split response payload, and the 60–90s latency on cold cache with expected <1s with Redis. This gives the agent realistic expectations for execution time and response shape.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is dense but well-structured, starting with the one-line purpose, then composition, limitations, and latency warning. Every paragraph earns its place, and formatting with bold emphasizes the key caveat sections without unnecessary filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the complexity of the federated operation, the description covers data sources, filtering boundaries, internal logic, latency, and output payload indications. With an output schema present and annotations marking the operation as safe and idempotent, nothing essential is missing for an agent to invoke the tool correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, but the description adds meaningful parameter semantics: max_resultados controls the most recent deduplicated results, and codigo_catmat/codigo_catser only filter the Dados Abertos leg while PNCP runs best-effort by modality. This adds value beyond the schema's field descriptions.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a clear statement that the tool federates Dados Abertos and PNCP to find similar contratações, which is a specific verb and resource. It further clarifies the composition by mentioning item-level querying, PNCP deduplication, and returning the most recent max_resultados, making it easily distinguishable from sibling tools.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description states the intended use case ('Insumo para mapear benchmarks de outros órgãos') and gives important limitations, such as PNCP not accepting catalog-item filters and CATMAT/CATSER only applying to the Dados Abertos side. It does not explicitly name alternative tools or exclusion conditions, but the context is clear enough for an agent to decide when to call it.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All MCP Compras.gov.br toolsInspect
Try in Browser
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

TDQS

A3.6/5.0
Disambiguation3/5

Most tools target distinct resources, and descriptions are extremely detailed, often explicitly warning about look-alikes. However, there is real overlap between composite and single-purpose tools (e.g., compras_checar_sancoes_fornecedor vs compras_perfil_fornecedor_completo vs compras_sancao_*), and similar-looking pairs like compras_contratos_consultar vs compras_contrato_comprasnet_consultar or compras_arp_listar vs compras_pncp_atas_listar require careful reading. With 100 tools, an agent will still face meaningful selection ambiguity.

Naming Consistency3/5

The dominant pattern is snake_case with a compras_ prefix, but the order and style vary: some are domain-first (compras_catmat_buscar), some are verb-first (compras_buscar_contratacoes_similares), and some are bare entity names with no verb (compras_sancao_ceis, compras_pncp_modalidades). The many listar/consultar/buscar variants are readable, but the convention is not predictable enough for a 100-tool surface.

Tool Count1/5

100 tools is an extreme count for any MCP server, regardless of domain breadth. Even if each tool has a legitimate upstream endpoint, this volume will heavily tax context windows and make reliable tool selection harder. Many tools could be consolidated into parameterized families (e.g., contratos, sancoes, pncp resources).

Completeness4/5

The server covers the Brazilian procurement domain remarkably well: catalogs, ARPs, 14.133 contracts, legacy regime, price research, suppliers, sanctions, PGC/PCA, PNCP, and Comprasnet contract subresources. Minor gaps remain, such as listing a supplier's full contract history without specifying an órgão, and some upstream limitations are only papered over with client-side workarounds.

Resources