codebolt.rag
Retrieval Augmented Generation tools for creating and querying knowledge indexes.
Available Tools
rag_create_index
- Create a new RAG indexrag_add_documents
- Add documents to a RAG indexrag_query
- Query documents from RAG indexrag_delete_index
- Delete a RAG indexrag_update_document
- Update a document in RAG indexrag_list_indexes
- List all available RAG indexes
Sample Usage
// Create a new RAG index
const createResult = await codeboltMCP.executeTool(
"codebolt.rag",
"rag_create_index",
{
indexName: "project_docs",
description: "Project documentation index"
}
);
// Add documents to index
const addResult = await codeboltMCP.executeTool(
"codebolt.rag",
"rag_add_documents",
{
indexName: "project_docs",
documents: [
{
id: "doc1",
content: "This is the API documentation...",
metadata: { type: "api", version: "1.0" }
}
]
}
);
// Query the RAG index
const queryResult = await codeboltMCP.executeTool(
"codebolt.rag",
"rag_query",
{
indexName: "project_docs",
query: "How to authenticate users?",
topK: 3
}
);
// List all indexes
const listResult = await codeboltMCP.executeTool(
"codebolt.rag",
"rag_list_indexes",
{}
);
// Update a document
const updateResult = await codeboltMCP.executeTool(
"codebolt.rag",
"rag_update_document",
{
indexName: "project_docs",
documentId: "doc1",
content: "Updated API documentation..."
}
);
info
This functionality is similar to the rag API and provides knowledge retrieval through MCP interface.