Best for
- Individuals who want to manage multiple AI agents in one place
- Teams that need agent collaboration, scheduling, and project organization
- Developers and users who want to self-host an AI application
L
LobeHub
Community
Organize and operate your AI team.
- Category
- Other #213 of 230
- Stars
- ★ 83.3k Very popular
- Runtime
- Docker · Node.js
- Credentials
- API key / credential required
- License
- Other / unspecified
- Last commit
33FMRS · D
The README presents LobeHub as a platform for organizing, collaborating with, scheduling, and managing memory for multiple AI agents. Because the server manifest was not provided, its specific MCP tools, transport, client compatibility, and installation configuration cannot be verified.
Strongest · Maintenance 12/20 Weakest · Reliability 2/20
Why each score
- Reliability 2/20
- No server manifest, MCP implementation files, tool definitions, initialization flow, or tests covering key MCP paths are provided. MCP initialization, tool-list handshake, and declared behavior therefore cannot be verified; the README product description is insufficient evidence for a higher score.
- Security and permissions 6/20
- The README documents an OPENAI_API_KEY and references plugins, Function Calling, MCP-compatible plugins, personal memory, and configurable remote proxies, indicating credential, external-network, and data-handling risks. However, there is no evidence of least-privilege scopes, data boundaries, per-tool confirmation, sensitive-data handling, or dangerous-operation safeguards, so substantial points are deducted. No explicit red-line risk is visible in the supplied material.
- Maintenance 12/20
- The project is not archived, and the README shows GitHub Actions, releases, tests, contributors, and named maintainers, with substantial public issue activity. However, actual commit history, release policy, dependency-update process, security-response channel, and license-file contents are not supplied; metadata also reports NOASSERTION. This supports only a mid-range score.
- Documentation 8/20
- The README provides substantial product, self-hosting, environment-variable, Docker, plugin, and local-development information. It does not provide an MCP manifest, server launch command, client connection configuration, tool parameters, permission model, limits, cost information, or MCP troubleshooting guidance, so points are deducted.
- Setup experience 5/20
- The README describes Vercel, Docker, and local application deployment, but not an MCP client connection workflow. There is no server entry point, transport specification, client configuration example, or verifiable evidence of a working MCP connection, resulting in a low setup score.
Static review · not runListed 2026-07-29
Read the FMRS scoring method →Fit and risk
What it can accessUses the networkControls a browser
Not for
- Users seeking a standalone MCP server with documented tools and transport
- Users who do not want to configure a model API key or deployment infrastructure
- Users requiring a plugin system that is described as fully stable and complete
Required permissions
- Self-hosting requires OPENAI_API_KEY
- Specific MCP tool permissions, client permissions, and authentication methods are not documented in the provided material
- Available agent plugins and skills depend on configuration
Risks and side effects
- The project is under active development, so functionality may change
- The plugin system is undergoing major development
- Personal memory can affect agent behavior and should be reviewed and managed
- Deployment and use may incur model API costs
Setup
Before you start
Runtime：Docker · Node.js
OPENAI_API_KEY
Required; create it at platform.openai.com/api-keys for model access.
Other optional settings (2)
OPENAI_PROXY_URL
Optional; overrides the default OpenAI API base URL (default https://api.openai.com/v1).
OPENAI_MODEL_LIST
Optional; controls the model list with + to add, - to hide, and model_name=display_name to rename.
The server manifest was not provided, so the standalone MCP server installation method, transport, and tools cannot be verified. The README states that the project can be self-hosted through Vercel, Zeabur, Sealos, Alibaba Cloud, or Docker. Self-hosting requires OPENAI_API_KEY; the Docker instructions include creating a storage directory, running setup.sh, and executing docker compose up -d.
Check that it works
Initialize with `bash <(curl -fsSL https://lobe.li/setup.sh)` and start with `docker compose up -d`; once the LobeHub web UI opens and you can chat using your OpenAI API key, the deployment works. Note: no specific MCP tool list is recorded for this repo; confirm available plugins and skills inside the app.
Troubleshooting
- Confirm that the required OPENAI_API_KEY is configured
- Check whether OPENAI_PROXY_URL overrides the default API endpoint as intended
- Check OPENAI_MODEL_LIST for correct model availability and display configuration
- For Docker deployment, confirm that setup.sh was run and docker compose up -d was executed
- For local development, confirm pnpm install completed and pnpm dev is running
Things to try
Once connected, you can ask your AI assistant things like:
- Create an agent that summarizes the news every morning and schedule it to run automatically
- Describe what I need to the Agent Builder and auto-configure a data analysis agent
- Use Pages to have multiple agents collaborate on polishing this article
- Connect my agent to the IM gateway so I can use it from my chat app
Use cases
Create and configure personalized AI agents
Coordinate multiple agents in parallel
Schedule agent runs
Organize agent work by project and workspace
Use structured, editable personal memory
Overview
LobeHub is an AI agent workspace that organizes agents into teams, schedules tasks, supports collaboration, and provides reporting. It includes an Agent Builder, Agent Groups, Pages, Schedule, Project, Workspace, editable Personal Memory, and self-hosting options.
Similar servers
Hevy MCP
77 · B
Manage your Hevy workout data from AI assistants via MCP.
Firecrawl MCP Server
75 · B
Firecrawl's official MCP server for web search, scraping, and structured extraction for AI agents
BioMCP
75 · B
One binary. One grammar. Evidence from the biomedical sources you already trust.
Compartment
75 · B
Encrypted, fully offline long-term memory for AI agents
Source revision 45b4a150b5c3 Data synced 2026-10-19 Read the FMRS scoring method