Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
RAG_MCP_RRF_KNoReciprocal Rank Fusion constant.60
RAG_MCP_TOP_KNoCandidates pulled from vector search before fusion.5
RAG_MCP_MAX_DEPTHNoDirectory-scan depth limit.20
RAG_MCP_CHUNK_SIZENoCharacters per chunk before overlap.700
RAG_MCP_MAX_TOKENSNoDirectory-scan size limit.2000000
RAG_MCP_EMBED_MODELNoEmbedding model name, meaning depends on provider.all-MiniLM-L6-v2
RAG_MCP_RERANK_TOP_KNoResults kept after reranking.5
RAG_MCP_CHUNK_OVERLAPNoCharacters shared between adjacent chunks.100
RAG_MCP_EMBED_API_KEYNoAPI key, only used by openai_compatible provider.
RAG_MCP_RERANKER_TYPENoReranker type: cross-encoder or disabled.cross-encoder
RAG_MCP_EMBED_BASE_URLNoOverride endpoint for openai_compatible provider. Defaults to official OpenAI endpoint.
RAG_MCP_EMBED_PROVIDERNoEmbedding provider to use: sentencetransformer (local), ollama (local server), or openai_compatible (remote).sentencetransformer
RAG_MCP_RERANKER_MODELNoReranker model, used when type is cross-encoder.cross-encoder/ms-marco-MiniLM-L-6-v2
RAG_MCP_WORKSPACE_ROOTNoThe root directory to search by default. Can be overridden with the doc_path parameter. If unset, the server uses the current working directory.

Capabilities

Server capabilities have not been inspected yet.

Tools

Functions exposed to the LLM to take actions

NameDescription

No tools

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/MasihMoafi/rag-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server