RAG-Anything MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| NEO4J_URI | No | Neo4j URI (default: bolt://localhost:7687). | bolt://localhost:7687 |
| MILVUS_URI | No | Milvus URI (default: rag_storage/milvus_lite.db). | rag_storage/milvus_lite.db |
| WORKING_DIR | No | Directory for RAG-Anything working files (no default, must be set). | |
| MINERU_TABLE | No | Enable table detection in MinerU (default: false). | false |
| MILVUS_DB_NAME | No | Milvus database name (default: empty). | |
| MINERU_API_KEY | No | API key for MinerU API (required when MINERU_EXECUTION_MODE=api). | |
| MINERU_FORMULA | No | Enable formula detection in MinerU (default: false). | false |
| NEO4J_DATABASE | No | Neo4j database name (default: neo4j). | neo4j |
| NEO4J_PASSWORD | No | Neo4j password (required when using Neo4JStorage). | |
| NEO4J_USERNAME | No | Neo4j username (default: neo4j). | neo4j |
| PARSER_BACKEND | No | MinerU parser backend (default: pipeline). | pipeline |
| NEO4J_WORKSPACE | No | Neo4j workspace name (default: raganything). | raganything |
| MILVUS_INDEX_TYPE | No | Milvus index type (default: AUTOINDEX). | AUTOINDEX |
| MILVUS_METRIC_TYPE | No | Milvus metric type (default: COSINE). | COSINE |
| EMBEDDING_BATCH_NUM | No | Number of items to embed in one batch (default: 4). | 4 |
| JOB_TIMEOUT_SECONDS | No | Overall timeout for a background job in seconds (default: 3600). | 3600 |
| MINERU_API_BASE_URL | No | Base URL for MinerU API (default: https://mineru.net). | https://mineru.net |
| MINERU_EXECUTION_MODE | No | MinerU execution mode: local or api (default: local). | local |
| MODEL_DOWNLOAD_POLICY | No | Model download policy: auto, never, or require-ready (default: auto). | auto |
| LIGHTRAG_GRAPH_STORAGE | No | Graph storage backend for LightRAG (default: NetworkXStorage). | NetworkXStorage |
| PARSER_TIMEOUT_SECONDS | No | Timeout in seconds for parsing a file (default: 120). | 120 |
| LIGHTRAG_VECTOR_STORAGE | No | Vector storage backend for LightRAG (default: NanoVectorDBStorage). | NanoVectorDBStorage |
| EMBEDDING_FUNC_MAX_ASYNC | No | Maximum number of concurrent embedding requests (default: 2). | 2 |
| EMBEDDING_TIMEOUT_SECONDS | No | Timeout for embedding requests in seconds (default: 180). | 180 |
| MINERU_API_TIMEOUT_SECONDS | No | Timeout for MinerU API requests in seconds (default: 1800). | 1800 |
| MULTIMODAL_TIMEOUT_SECONDS | No | Timeout for multimodal postprocessing in seconds (default: empty, meaning no limit). | |
| TEXT_INGEST_TIMEOUT_SECONDS | No | Timeout in seconds for text ingestion phase (default: 1800). | 1800 |
| MODEL_DOWNLOAD_TIMEOUT_SECONDS | No | Timeout in seconds for model downloads (default: 1800). | 1800 |
| MINERU_API_POLL_INTERVAL_SECONDS | No | Polling interval in seconds for MinerU API (default: 5). | 5 |
Capabilities
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
No tools | |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
Latest Blog Posts
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
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/yyisahandsomeboy/rag_anything_mcp_sever'
If you have feedback or need assistance with the MCP directory API, please join our Discord server