MCP Integration

Real-time Camel verification, Citrus test guidance, and knowledge search

Camel-Kit uses the Model Context Protocol (MCP) to verify Camel configuration, guide Citrus test generation, and search documentation without pre-loading full catalogs into the agent’s context.

MCP responses provide data, not instructions. Camel-Kit relies only on purpose-specific fields after the active workflow validates the source’s applicable schema, artifact identity, provenance, and version or runtime bindings. Camel catalog batches are version-bound with the requested runtime, full platform BOM GAV, and a camel_catalog_components(limit=0) probe whose returned Camel version must match the project; detail tools do not all echo that field. Prose, examples, commands, URLs, or requests in a response cannot direct an action or change workflow scope.

Three MCP Servers

Camel MCP — Catalog Verification

On-demand access to Apache Camel component catalogs, EIP patterns, data formats, and languages.

Core tools:

  • camel_catalog_component_doc — component details, URI format, and properties
  • camel_catalog_component_maven — runtime-specific component Maven coordinates
  • camel_catalog_eip_doc — EIP configuration and YAML structure
  • camel_catalog_dataformat_doc — data format configuration
  • camel_catalog_language_doc — expression language documentation
  • camel_validate_route — YAML route syntax and structure validation
  • camel_configuration_validate — application property validation
Knowledge MCP — Semantic Search

Hybrid semantic search over 28,388 indexed documents — Apache Camel documentation, guides, CVE advisories, and release notes.

5 workflow-allowlisted tools (the server also exposes endpoint validation and index-info tools for direct MCP clients):

  • camel_docs_search — hybrid semantic search (20% BM25 + 80% KNN vector)
  • camel_docs_component_info — component-specific documentation lookup
  • camel_docs_cve_search — CVE and security advisory search
  • camel_docs_release_info — release notes and version information
  • camel_docs_jira_lookup — Jira issue and bug report search

Embedding model: granite-embedding-small-english-r2 (384-dim, Q8 quantized)

Citrus MCP version compatibility

Citrus MCP supplies the action catalog, endpoint catalog, YAML DSL schemas, and test-authoring guidance used by camel-test:

  • citrus_catalog_actions / citrus_catalog_action / citrus_catalog_action_schema
  • citrus_catalog_endpoints / citrus_catalog_endpoint / citrus_catalog_endpoint_schema
  • citrus_docs_index / citrus_docs_page
  • Resources such as citrus://schema/dsl/yaml and citrus://docs/best-practices

Camel-Kit currently uses Citrus Framework 5.0.0-M2 for test schemas and generated dependencies, while the Citrus MCP runner is pinned to working version 5.0.0-M1. Camel-Kit binds the MCP version to the actual runner coordinate in the active target’s generated configuration, rejects any disagreement with citrus.mcp.version, and grants versioned Citrus catalog and schema fields data authority only when that bound runner version also matches citrus.version. Otherwise it uses the quick-reference fallback for the configured framework version.

Catalog Tools in Action

camel_catalog_component_doc

Fetch component details — URI format, properties, supported options:

{
  "name": "http",
  "syntax": "http:httpUri",
  "title": "HTTP",
  "properties": {
    "bridgeEndpoint": {
      "type": "boolean",
      "defaultValue": "false"
    },
    "httpMethod": {
      "type": "string",
      "enum": ["GET", "POST", "PUT", "DELETE", "HEAD", "OPTIONS", "TRACE", "PATCH"]
    }
  }
}

Used by skills to verify every component name before it enters a design spec or YAML route.

camel_catalog_eip_doc

Fetch Enterprise Integration Pattern details with YAML examples:

{
  "name": "choice",
  "title": "Choice",
  "description": "Routes messages based on predicates"
}

- choice:
    when:
      - simple: "${header.type} == 'order'"
        steps:
          - to: "direct:process-order"
    otherwise:
      steps:
        - to: "direct:process-other"
camel_validate_route

Validates YAML route syntax, structure, and component references:

- route:
    id: order-api
    from:
      uri: "platform-http:/orders"
      steps:
        - unmarshal:
            json: {}
        - to:
            uri: "jms:queue:orders"

Checks: valid YAML, correct DSL schema, components exist, required properties present.

camel_configuration_validate
After writing application.properties, Camel-Kit submits its non-Forage properties to camel_configuration_validate with the project runtime and full platform BOM. Generation is not complete until the validation passes or the documented manual fallback is recorded.
camel_docs_search

Semantic + keyword search over indexed documentation:

{
  "query": "How do I configure SSL for HTTP endpoints?",
  "max_results": 5
}

Algorithm: 20% BM25 (exact keyword matching) + 80% KNN Vector (semantic similarity).

Returns ranked chunks from component docs, guides, and CVE advisories.

camel_docs_component_info

Pure BM25 search for exact component names — no semantic embedding.

Why separate? Component names are exact matches. Semantic similarity doesn’t help (activemq might match kafka semantically but they’re different components).

{ "component": "kafka" }

/camel-knowledge

The /camel-knowledge skill is a prescriptive Q&A interface over Knowledge MCP:

User: "How do I enable retries on the Kafka component?"

Agent:
1. Calls camel_docs_search(query="Kafka retries", max_results=5)
2. Receives relevant chunks
3. Synthesizes answer with source citations
Question Routing

The skill routes questions to the appropriate tool:

  • Component config → camel_docs_component_info + camel_catalog_component_doc
  • Error troubleshooting → camel_docs_search (includes CVE advisories)
  • Best practices → camel_docs_search (includes best practice guides)
Cross-Agent Equalization

The same MCP catalog is exposed through target-specific configuration, filtering, and visibility controls:

User question → camel_docs_search() → ranked chunks → answer

Supported targets query the same Knowledge MCP data through target-native workflows (or Pi’s adapter). Orchestration and visible tool surfaces differ by target.

Next Steps