LobeHub — Chief Agent Operator for Multi-Agent Work
Organize your AI agents into 7×24 operations by hiring, scheduling, and reporting on your entire AI team.
- Source repo
- lobehub/lobehub
- Stars
- ★ 83k
- Last updated
- today
- License
- NOASSERTION
- Primary language
- TypeScript
- FA score
- 48/100 · Major gaps
At a glance
- How it runs
- Works with
- Universal · cross-platformChatGPT · OpenAI API · Claude API
- Cost
- Free software; you pay for model usage
- Setup effort
- Medium · a few setup steps
- You'll need
- Typical use
- A developer wants to unify access to multiple models (GPT, Claude, DeepSeek) and manage them as a cohesive team.
- Not a fit if
- Users who want zero-config setup without Docker or env vars
- Teams needing stability, not a fast-changing pre-release product
- Source review
- 48/100 · Major gaps
What does this agent do, and when should you use it?
LobeHub is a work-and-lifestyle space for finding, building, and collaborating with agent teammates that grow with you. It treats agents as the unit of work, providing infrastructure where humans and agents co-evolve. Key features include an Agent Builder that auto-configures agents from descriptions, Agent Groups for parallel collaboration, Personal Memory for continual learning and transparency, and an Operator mode that hires, schedules, and reports on your entire AI team. The platform supports self-hosting via Docker or one-click deployment on Vercel, Zeabur, Sealos, RepoCloud, and Alibaba Cloud. It offers a plugin ecosystem with 10,000+ skills and MCP compatibility. The LobeHub ecosystem includes libraries like @lobehub/ui, @lobehub/icons, @lobehub/tts, and @lobehub/lint.
LobeHub enables users to create and orchestrate AI agents as a team. The Agent Builder allows describing a need and automatically configuring an agent for immediate use. Agent Groups assemble the right agents for tasks, enabling parallel collaboration and iterative improvement. Pages allow multiple agents to write/refine content with a shared context. Schedule runs agents at specified times, even when the user is offline. Project and Workspace organize work by project and provide shared team spaces. Personal Memory builds a model of user needs, with white-box memory that users can inspect and edit. The system supports connecting to any model via an IM Gateway and any modality, with unified intelligence. Users can extend functionality through plugins using Function Calling, and the platform provides a plugin gateway (Edge Function) for backend services. Deployment options include Docker (setup script and docker compose up -d), Vercel, Zeabur, Sealos, RepoCloud, and Alibaba Cloud, with required environment variable OPENAI_API_KEY and optional OPENAI_PROXY_URL and OPENAI_MODEL_LIST.
- A developer wants to unify access to multiple models (GPT, Claude, DeepSeek) and manage them as a cohesive team.
- An operations manager needs agents to run scheduled tasks (e.g., generate reports, process data) while they are offline.
- An individual who wants a personalized AI assistant that continually learns their work habits and preferences.
- A technical user who wants to extend agent capabilities with MCP-compatible plugins to integrate with daily tools like documents and databases.
- A content creation team that needs multiple agents to collaborate on the same document (e.g., editing, proofreading, polishing) with shared context.
How do you install or deploy this agent?
Installation options: 1. One-click deployment on Vercel, Zeabur, Sealos, or Alibaba Cloud (requires OpenAI API Key). 2. Self-host with Docker: create a folder mkdir lobehub-db && cd lobehub-db, run bash <(curl -fsSL https://lobe.li/setup.sh), then docker compose up -d. Set environment variable OPENAI_API_KEY (required), optionally OPENAI_PROXY_URL and OPENAI_MODEL_LIST.
How do you use this agent?
Access the web interface after deployment. Use the Agent Builder to describe your need and get an auto-configured agent. Assemble agents into Agent Groups for parallel tasks, use Pages for shared-context writing, Schedule for time-based runs, and Project/Workspace for organization. Configure Personal Memory to let agents learn your habits. Connect to 10,000+ skills via the plugin marketplace or MCP-compatible plugins, and extend capabilities with Function Calling.
What are this agent's strengths and limitations?
- Unified interface for managing multiple AI agents, reducing tool fragmentation.
- Agent Groups and scheduling enable automated multi-agent workflows that run 24/7.
- Rich ecosystem: 10,000+ skills, MCP support, and development libraries (UI, Icons, TTS, Lint).
- Flexible self-hosting options (Docker, Vercel, Alibaba Cloud) lower the barrier for private deployment.
- Actively under development, features may change frequently; user should expect bugs.
- Requires OpenAI API Key for default operation; using other models may need additional proxy configuration.
- Self-hosting configuration is non-trivial (environment variables, Docker setup) for non-experts.
- Complex multi-agent collaboration workflows may have a steep learning curve.
How does this agent compare with similar options?
The platform positions itself as a Chief Agent Operator, distinct from single-model chat tools like ChatGPT or Claude. It competes more with agent orchestration frameworks like LangChain or AutoGen, but LobeHub emphasizes a productized, out-of-the-box experience with operational management features.
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| LobeHub — Chief Agent Operator for Multi-Agent Work This agent | 48 · Major gaps | Self-hosted serviceFree + model costs | ★ 83k | today | TypeScript | ChatGPT · OpenAI API · Claude API |
| AGNT Agent OS | 52 · Major gaps | Desktop appFree + model costs | ★ 554 | today | JavaScript | Codex · Claude Code · OpenAI API · Claude API |
| EpicStaff Agent Workflow Platform | 57 · Major gaps | Self-hosted service | ★ 359 | 10d ago | Python | — |
| Auto Company | 70 · Some gaps | CLIFree + model costs | ★ 3.1k | 1d ago | Python | Codex · Claude Code |
How does FollowAgents rate this agent?
Why each dimension lost points
Evidence shows the project provides a security policy (SECURITY.md) with supported versions and vulnerability reporting, but lacks specific implementation details on least privilege, user confirmation, data flow transparency. Dependency security: package.json lists many dependencies but no explicit audit or vulnerability scanning evidence. External effects: README mentions deployment and plugins but not permission controls. Rollback: release process exists but no explicit rollback mechanism. Source attribution: author and repo info present, but publisher unverified. Thus scores are low due to insufficient evidence.
Project has test files (e.g., file-loaders tests) indicating some reliability, but no full test coverage or failure message design. Dependency availability: package.json lists many dependencies but no availability guarantees. Failure messages: no explicit error handling or user prompts seen in source. Thus scores are moderate-low.
README describes multiple use cases (deployment, plugin development) but lacks clear capability boundaries. Trigger precision: no explicit trigger mechanism description. Environment fit: provides multiple deployment options (Vercel, Docker) but not detailed environment requirements. Thus scores are moderate.
Information architecture is clear with README, docs links, and directory structure. Install notes are detailed with multiple deployment methods. Naming stability: version numbers and release process exist. Examples and FAQ: README has examples but no FAQ. Known limitations: README mentions active development but no specific limitations. License: LICENSE file is custom LobeHub Community License, but package.json declares MIT, inconsistency. Versioning changelog: CHANGELOG generation process exists. Maintenance responsibility: principal maintainers listed. Thus scores are moderate.
Output usability: README describes features but no concrete output examples. Marginal value: project offers unique multi-agent collaboration features but no comparison with alternatives. Cost benefit: no cost information. Thus scores are moderate-low.
Claim traceability: feature descriptions in README lack implementation evidence. Cross-source corroboration: no external validation. Fact-inference separation: not clearly distinguished. Thus scores are low.
- License inconsistency: LICENSE file is LobeHub Community License, but package.json declares MIT, needs clarification.
- Publisher unverified, identity unknown, handle with caution.
- Many dependencies without evidence of audit or vulnerability scanning, need to check yourself.
- Feature descriptions in README lack implementation details, need further verification.
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
🐤 Canary Build — v2.2.21-canary.2 > Automated canary build from canary branch. Commit Information - Based on changes since v2.2.21-canary.1 - Commit count: 82 - 9b583f3b1e 🐛 fix: preserve environment variable raw editor values (#21044) (AmAzing-) - fd940ffd24 👷 ci(release): fix README sync on tag checkout and retry macOS staple (#21155) (Lobehub Coder) - 465a2220f9 🐛 fix: keep environment details usable in narrow windows (#21117) (AmAzing-) - 53994223eb 🐛 fix: distinguish GitHub repository errors from empty results (#21174) (AmAzing-) …View on GitHub ↗ -
v2.2.20 🚀 Release v2.2.20
Release notes excerpt
📦 Release v2.2.20 This release was automatically published from PR #21078. Changes See PR description: https://github.com/lobehub/lobehub/pull/21078 Commit Message 🚀 LobeHub Release (20261018) Release Date: October 18, 2026 Since v2.2.19: 419 merged PRs · 14 contributors > Agents now reach further across your devices, and the app opens faster. Topics, documents and device data load from local copies first, and server runs can use models that only exist on your device. --- ✨ Highlights …View on GitHub ↗