arc53/DocsGPT

Private AI platform for agents, assistants and enterprise search. Built-in Agent Builder, Deep research, Document analysis, Multi-model support, and API connectivity for agents.

DocsGPT – AI‑powered agents that answer from your documents

What it is

  • An open‑source platform that lets you turn any collection of files, web pages or connected apps (Google Drive, Confluence, GitHub, etc.) into a private AI agent. The agent can be queried via a web UI, a chat widget, an OpenAI‑compatible API, or integrations such as Discord, Slack and Telegram.
  • All components run in your own environment – Docker, Kubernetes, pip install, or even an air‑gapped machine – so you keep the data and the model under your control.

Core capabilities

Feature What it does
Agents & research Define an agent with its own prompt, knowledge base, tools and model. A research mode lets the agent perform multi‑step reasoning and cite sources.
Visual workflows Drag‑and‑drop builder (React Flow) to chain agents, add conditional logic, maintain state, and run sandboxed code.
Knowledge ingestion Connectors for PDFs, Office files, audio, web pages and many SaaS sources (Google Drive, SharePoint, Confluence, GitHub, S3, …). Supports continuous sync and a GraphRAG pipeline.
Cited answers Every response includes links to the exact document fragments that were used.
Tool integration Built‑in web‑search, arbitrary REST‑API calls, MCP‑server access, sandboxed code execution, and remote shell on paired devices.
Model agnostic Works with any OpenAI‑compatible endpoint: OpenAI, Anthropic, Google, Groq, OpenRouter, or local models via Ollama, vLLM, etc.
Guardrails Automatic detection and redaction of PII, secrets, prompt‑injection attempts, and ungrounded answers.
Scheduling & webhooks Run agents on a timer or trigger them from external systems via HTTP.
Team features SSO (OIDC/SCIM), role‑based access control, usage quotas, audit logs, analytics and OpenTelemetry support.

How you run it

  • One‑command installer (macOS/Linux): curl -fsSL https://docs.ac/install | bash – pulls Docker, installs the docsgpt CLI and starts the stack on http://localhost:7091.
  • Alternative deployments: Docker Compose, pip install, Kubernetes manifests, or manual source checkout.
  • CLI commands (docsgpt status, logs, upgrade, backup, down) let you manage the service.

How you can use it

  • Embed a searchable support bot on any website via a script tag or the npm docsgpt React widget.
  • Call the OpenAI‑compatible endpoint (/v1) from any existing codebase, or use the dedicated Agent API and webhook system.
  • Connect to MCP clients (Claude, Cursor, etc.) so they can search your private knowledge.
  • Deploy bots to chat platforms (Discord, Slack, Telegram) using the community extensions.

Tech stack

  • Backend: Python, Flask + Starlette (uvicorn/gunicorn), Celery, Pydantic, PostgreSQL, Redis, vector stores (FAISS, pgvector, Elasticsearch, Qdrant, Milvus, MongoDB).
  • Frontend: React (Vite, Redux Toolkit, Tailwind CSS) with a visual workflow editor built on React Flow.
  • Docs site: Next.js + Nextra.

License

  • MIT License – you can freely use, modify and commercialize the software.

Where to learn more

DocsGPT is a genuine, self‑hostable AI‑agent framework focused on grounding LLM answers in your own documents, making it a solid fit for enterprises, developers, and anyone who needs private, source‑cited AI assistance.

Related

  • Project
    labring/FastGPT

    FastGPT is an open‑source platform for building AI agents and RAG‑enabled applications. It offers a visual workflow editor, built‑in knowledge‑base handling, plugin hot‑reload, debugging tools, and can be launched instantly with Docker or via a hosted cloud service.

  • Project
    Significant-Gravitas/AutoGPT

    AutoGPT is an AI-powered platform that lets you manage a team of specialized agents to automate recurring professional tasks like content creation, research, and administrative work.

  • Project
    langflow-ai/openrag

    An intelligent agent-powered document search platform that enables users to upload, process, and query documents through a chat interface using RAG.

  • Project
    the-open-agent/openagent

    An open-source personal AI assistant platform that combines LLMs, RAG, and autonomous agent loops into a single self-hostable binary.

  • Project
    CopilotKit/OpenBot

    OpenBot is an open‑source, self‑hosted platform that lets you run AI “coworker” agents inside your own infrastructure. Each Bot gets its own container with a browser and file workspace, all tool calls pass through a policy‑enforced gateway that logs every action. Built on Docker Compose, Bun, PostgreSQL and the AG‑UI protocol, it works with any LLM (OpenAI, Anthropic, Google, or self‑hosted) via CopilotKit Intelligence. Features include per‑Bot computers, CEL‑based access policies, human‑in‑the‑loop hand‑offs, component‑rich responses, audit trails, encrypted credentials, and scheduled routines. The repo is a template (MIT‑licensed, Alpha status) meant to be cloned and customized for enterprise AI assistants, RAG bots, or automated workflows.