MaxKB
A self-hosted platform for enterprise knowledge Q&A and agent workflow orchestration.
- Source repo
- 1Panel-dev/MaxKB
- Stars
- ★ 23k
- Last updated
- 1d ago
- License
- GPL-3.0
- Primary language
- Python
- FA score
- 24/100 · Major gaps
At a glance
- How it runs
- Works with
- Universal · cross-platformOpenAI API · Claude API
- Cost
- Free software; you pay for model usage
- Setup effort
- Medium · a few setup steps
- You'll need
- Typical use
- An enterprise knowledge-management team uploads internal documents to provide employees with RAG-based knowledge Q&A.
- Not a fit if
- Teams wanting a turnkey setup with no model or database configuration
- Production environments with strict rules against published default admin credentials
- Source review
- 24/100 · Major gaps 8 safety controls not found
What does this agent do, and when should you use it?
MaxKB, short for Max Knowledge Brain, is an open-source platform for building enterprise-grade agents with a Vue.js frontend, Python/Django backend, LangChain, and PostgreSQL plus pgvector. It brings document upload or online-document crawling, text splitting, vectorization, and RAG-based Q&A into one product. Its workflow engine, function library, and MCP tool-use support the orchestration of more complex AI processes. It supports private models such as DeepSeek, Llama, and Qwen, alongside public models including OpenAI, Claude, Gemini, and MiniMax. It is deployed as a Docker container with a web interface, making it relevant for teams that want to self-host knowledge bases and intelligent Q&A.
MaxKB accepts directly uploaded documents or automatically crawls online documents, then automatically splits and vectorizes text for its RAG Pipeline and intelligent Q&A. Its workflow engine combines a function library with MCP tool-use to orchestrate AI processes. The platform supports text, image, audio, and video input and output, and is described as enabling zero-code integration with third-party business systems. Deployment runs docker run -d --name=maxkb --restart=always -p 8080:8080 -v ~/.maxkb:/opt/maxkb 1panel/maxkb, after which the service is accessed through its web interface.
- An enterprise knowledge-management team uploads internal documents to provide employees with RAG-based knowledge Q&A.
- A customer-service team connects an existing business system to MaxKB to provide an intelligent Q&A entry point.
- A researcher crawls online material and incorporates it into a knowledge-Q&A flow for academic research support.
- An education organization uses a document knowledge base and workflows to organize intelligent Q&A for teaching.
- A team that needs private deployment of DeepSeek, Llama, or Qwen uses one platform to build enterprise agents.
How do you install or deploy this agent?
On a server that can run Docker, execute: docker run -d --name=maxkb --restart=always -p 8080:8080 -v ~/.maxkb:/opt/maxkb 1panel/maxkb. After the container starts, open http://your_server_ip:8080. The default administrator username is admin and the password is MaxKB@123... For Docker image pull failures in China, the README points to offline-installation documentation, but does not provide offline installation commands here.
How do you use this agent?
Open http://your_server_ip:8080 and sign in with admin / MaxKB@123... Upload documents or use automatic online-document crawling; MaxKB performs automatic text splitting and vectorization for RAG Q&A. For more involved processes, use the workflow engine with the function library and MCP tool-use. Its stated zero-code third-party-system integration can be used to add Q&A capability to an existing business system.
What are this agent's strengths and limitations?
- Combines document upload, online-document crawling, text splitting, vectorization, and RAG Q&A in one platform.
- Provides a workflow engine, function library, and MCP tool-use in addition to knowledge Q&A.
- Explicitly supports both private models and public models including OpenAI, Claude, Gemini, and MiniMax.
- Documents a single Docker startup command, service port, and persistent-volume mount path.
- Default administrator credentials are published in the deployment instructions, so production deployments need their own access-security controls.
- The README does not document provider-specific model configuration, authentication fields, or feature differences between models.
- Its documented stack includes PostgreSQL plus pgvector, but no process is described for migrating existing data to that combination.
- Offline installation is only linked as documentation; the README itself does not provide complete offline-install steps.
How does this agent compare with similar options?
MaxKB explicitly supports both private models, such as DeepSeek, Llama, and Qwen, and public models including OpenAI, Claude, Gemini, and MiniMax; the README does not compare it with other agent platforms.
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| MaxKB This agent | 24 · Major gaps | Self-hosted serviceFree + model costs | ★ 23k | 1d ago | Python | OpenAI API · Claude API |
| Dive into LangGraph | 69 · Some gaps | Library / SDKFree + model costs | ★ 468 | 1mo ago | Jupyter Notebook | Claude Code |
| AI Agent Engineering Textbook | 55 · Major gaps | FrameworkFree | ★ 651 | 13d ago | HTML | — |
| CUGA | 51 · Major gaps | CLIFree + model costs | ★ 1.1k | 2d ago | Python | OpenAI API |
How does FollowAgents rate this agent?
Why each dimension lost points
Evidence shows no documentation of permission model, user confirmation, data flow transparency, sensitive data handling, dependency security audit, external effect control, rollback, or source attribution. All criteria scored 0 due to lack of evidence.
Self-consistency: README claims multimodal support but pyproject.toml lacks audio/video libraries, minor inconsistency, scored 1. Dependency availability: dependency list is complete and pinned, but no source or mirror info, scored 1. Failure messages: no error handling or user-facing failure documentation, scored 0.
Audience and scenarios: README clearly lists customer service, knowledge base, research, education, scored 2. Capability boundaries: feature list clear but no limitations, scored 1. Trigger precision: no trigger mechanism or precise instructions, scored 0. Environment fit: Docker quick start and offline install docs, but no system requirements, scored 1.
Information architecture: README structured but lacks detailed docs links, scored 1. Install notes: Docker command and offline install docs, scored 2. Naming stability: project name and version clear, but no naming conventions, scored 1. Examples and FAQ: screenshots but no examples or FAQ, scored 1. Known limitations: not mentioned, scored 0. License: GPL-3.0 license file complete, scored 2. Versioning and changelog: version number but no changelog, scored 1. Maintenance responsibility: author info clear but no maintenance policy, scored 1.
Output usability: screenshots but no output format or integration examples, scored 1. Marginal value: feature-rich but no comparison with alternatives, scored 2. Cost benefit: no performance or resource data, scored 1.
Claim traceability: feature claims not linked to code or docs, scored 1. Cross-source corroboration: Docker image and badges, but no independent verification, scored 1. Fact-inference separation: no distinction between facts and inferences, scored 0.
- Not found in source: least-privilege scopingGrant only what the task needs: a dedicated account or read-only token, scoped to specific directories and repos.
- Not found in source: confirmation before actingTurn on (or add) a confirmation step before it acts, and try it in a sandbox or test environment before real data.
- Not found in source: data-flow disclosureWatch which external services it contacts (proxy or firewall logs) and keep sensitive data out until you know where it goes.
- Not found in source: sensitive-data handlingUse dedicated, low-privilege, revocable API keys — never production credentials — and keep secrets out of logs.
- Not found in source: dependency securityPin versions and run a dependency audit (npm audit, pip-audit) before installing; prefer running it in a container.
- Not found in source: disclosed external effectsEstablish which external systems it writes to, sends to or changes, and verify with test accounts or repos before production.
- Not found in source: rollback or recovery pathBack up first, or work on a git branch or snapshot, so its changes can be undone.
- Not found in source: verifiable attributionInstall from the official repo or registry and check the publisher and URL to avoid look-alike packages.
- Default admin credentials are hardcoded in README, a security risk; users should be prompted to change immediately.
- Dependencies include large libraries like torch, but no resource requirements or performance benchmarks are provided; assess before deployment.
- No permission model or data flow transparency documentation; conduct your own audit before enterprise deployment.
Recent releases
Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.
-
Release notes excerpt
Security Fixes ■ Security (X-Pack): Fixed a vulnerability where the login authentication configuration endpoint did not redact sensitive fields, exposing third-party login app credentials (such as DingTalk app_key, corp_id, and app_secret) in plaintext, which could lead to enterprise app takeover. Bug Fixes ■ Knowledge Base: Fixed an issue where special characters such as \, /, :, *, and ? could be entered when quickly creating a blank document. …View on GitHub ↗ -
Release notes excerpt
Enhancements - Knowledge Base: Special characters such as \, /, ?, *, and [] are now disallowed when renaming documents. - Agent: A confirmation prompt is now shown when deleting conversation history, to prevent accidental deletion. - Q&A Page: Tables in AI responses now support horizontal scrolling when the number of columns exceeds the visible width. Bug Fixes - Agent: Fixed an issue where content wrapped in html_render tags was rendered blank when exported to PDF or image (#7222). …View on GitHub ↗ -
Release notes excerpt
Security Vulnerability Fixes - Security: Fixed a security vulnerability where Chat file uploads could spoof APPLICATION attribution and be read anonymously, allowing content injection into the application namespace. - Security: Fixed a security vulnerability where document/paragraph migration only validated source knowledge base permissions, leading to cross-tenant content injection and RAG poisoning. …View on GitHub ↗
FAQ
What license does MaxKB use?
Which models can it use?
What are the first-login credentials?
admin and password MaxKB@123...What is required for the documented deployment path?
~/.maxkb.