AnythingLLM

Container running AnythingLLM - an open-source, all-in-one RAG (Retrieval-Augmented Generation) and AI platform.

Usage with docker

You may start Ollama before if you want to use a local model.

• Start AnythingLLM: docker compose up -d
• Open https://anythingllm.dev.localhost/
• Follow instructions

Add MCP servers

See sample config plugins/anythingllm_mcp_servers.json and use the following to copy it to /app/server/storage/plugins/anythingllm_mcp_servers.json in the container :

docker cp plugins/anythingllm_mcp_servers.json anythingllm:/app/server/storage/plugins/anythingllm_mcp_servers.json

# restart or refresh in https://anythingllm.dev.localhost/settings/agents
docker restart anythingllm

Resources

• Mintplex-Labs/anything-llm
• AnythingLLM Documentation
• AnythingLLM Docker Hub