FollowAgents AI AGENTS DIRECTORY
Home / All Agents / MetaBot Personal Agent Workspace
Dev & Engineering codex-clikimi-codefeishu-bottelegram-botagent-teamstask-schedulermetamemory

MetaBot Personal Agent Workspace

Supervise multi-engine coding agents from chat apps or a local web console.

View on GitHub ↗ Quick start ↓
Source repo
xvirobotics/metabot
Stars
★ 1k
Last updated
13d ago
License
MIT
Primary language
TypeScript
FA score
59/100 · Major gaps

At a glance

How it runs
Self-hosted serviceCLIWeb app
Works with
Universal · cross-platformCodex · Claude CodeOpenAI API · Claude API (Partial support)
Cost
Free software; you pay for model usage
Setup effort
Medium · a few setup steps
You'll need
Node.js >= 22.19GitCodex CLI, Kimi Code, or a Claude-compatible CLI setupchat-channel credentialsShell / CLINetwork accessLocal filesystem
Typical use
A solo developer away from their desk sends a coding task to a logged-in Codex bot through Feishu or Telegram, then checks test execution and generated output from the phone.
Not a fit if
  • Users who don't want to run local services and manage tokens/ports
  • Teams wanting a zero-dependency hosted service
  • Anyone treating the WeChat channel as production-ready
Source review
59/100 · Major gaps

What does this agent do, and when should you use it?

MetaBot is a self-hosted personal agent workspace composed of a local Core, Core Console, IM Bridge, and the metabot CLI. It routes messages from Feishu/Lark, Telegram, WeChat, or local Web Chat to configured bots, which execute in their own workspaces through Codex CLI, Kimi Code, or a Claude Code-compatible path. Each bot can have separate channel credentials, engine, working directory, and session while collaborating through Agent Teams and Agent Bus. The Core Console combines Chat, Agents, Memory, Skills, T5T, Teams, CLI Access, and diagnostics behind a local Bearer token. It fits individual developers prepared to run local services and grant agents read, write, and code-execution access within configured workspaces.

The Bridge receives Feishu/Lark WebSocket traffic, Telegram or WeChat long-polling traffic, or Web Chat input, then routes it according to bot configuration. Its Codex path runs codex exec --json and codex exec resume, producing JSONL streaming output, tool activity, and resumed sessions; its Kimi Code path uses the official local Server API for persistent sessions, atomic snapshots, questions, stop/resume, tools, sub-agents, and Goals. The console displays streamed replies, tool execution, and output files, and lets an operator answer agent questions or stop a run. MetaBot also provides cross-session MetaMemory retrieval, persistent Agent Teams tasks and run state, multi-turn /goal checkpoints, and supported /background tasks.

  1. A solo developer away from their desk sends a coding task to a logged-in Codex bot through Feishu or Telegram, then checks test execution and generated output from the phone.
  2. An engineer maintaining separate repositories assigns distinct working directories to a Codex development bot and a Kimi review bot, then coordinates them through Agent Teams.
  3. A developer supervising local code changes uses Core Console to inspect streamed responses, tool calls, and output files, and stops a run when necessary.
  4. A project owner who needs context across conversations uses MetaMemory to retrieve knowledge and can synchronize it to a Feishu knowledge base.
  5. A Feishu group owner operating several development bots uses @Bot /group-reply mention|all|status to set each bot's group response mode.

How do you install or deploy this agent?

Requirements: Node.js >= 22.19, Git, at least one engine, and credentials for a chat channel. Run:

curl -fsSL https://github.com/xvirobotics/metabot/releases/latest/download/install.sh | bash

Install and log in to an engine in a separate terminal, for example:

npm install -g @openai/codex
codex login

Or:

npm install -g @moonshot-ai/kimi-code@latest
kimi login

Follow the setup flow for Feishu/Lark, Telegram, or WeChat, then run metabot status and metabot health. Open http://localhost:9200, enter the token in ~/.metabot-core/token, and choose a bot.

How do you use this agent?

Configure each bot in bots.json with its name, engine, channel credentials, and defaultWorkingDirectory; codex is the default when engine is omitted. Send work through a configured chat channel or http://localhost:9200. Use /model to inspect or switch the current engine/model, /effort low|medium|high|xhigh|max|ultra to set Codex reasoning effort, /status for session information, and /stop to halt a task. Use /goal <condition> for supervised multi-turn work and /background <prompt> for supported background tasks.

What are this agent's strengths and limitations?

Pros
  • A single Bridge process can mix Codex, Kimi Code, and Claude Code-compatible bots while keeping each bot's workspace, session, and channel credentials distinct.
  • Its documented Feishu/Lark WebSocket and Telegram/WeChat long-polling modes do not require opening a public inbound port.
  • Core Console puts chat, tool output, stop controls, Memory, Skills, T5T, Teams, and diagnostics behind one token-authenticated entry point.
  • The installer verifies release checksums, and updates verify SHA256SUMS, the complete personal-edition manifest, and version data.
Limitations
  • It requires Node.js >= 22.19, Git, an external engine, and chat-channel credentials, so it is not a zero-dependency setup.
  • The public Codex adapter currently uses codex exec; Codex app-server support and steering during Codex/Kimi Feishu execution are described as future work.
  • The WeChat integration is labeled as limited beta, so it should not be treated as a mature channel guarantee.
  • Agents can read, write, and execute code in configured workspaces; operators must protect tokens, restrict IM bot visibility, and secure exposed ports with their own authenticated proxy or private network.

How does this agent compare with similar options?

Engine choice: Codex is the default, Kimi Code is a first-class optional engine, and a Claude Code-compatible path is retained for existing Claude bots, sessions, skills, and workspaces.

Key facts side by side with the most closely related agents.

Agent Source review Form / cost Stars Updated Language Full support on
MetaBot Personal Agent Workspace This agent 59 · Major gaps Self-hosted serviceFree + model costs ★ 1k 13d ago TypeScript Codex · Claude Code
Codexia 57 · Major gaps Desktop appFree + model costs ★ 941 17d ago TypeScript Codex · Claude Code
Oh My Codex (OMX) 67 · Some gaps CLIFree + model costs ★ 34k today TypeScript Codex
Claude-to-IM Bridge 55 · Major gaps Agent plugin / skillFree + model costs ★ 2.9k 7mo ago TypeScript Codex · Claude Code

How does FollowAgents rate this agent?

FollowAgents source review · FARS-2.1
Major gaps
59/ 100 5-point scale 3.0 / 5
Trust 14/29
Reliability 8/14
Adaptability 12/18
Convention 13/18
Effectiveness 7/13
Verifiability 5/8
Why each dimension lost points
Trust14 / 29 · 2.4/5

Security docs specify workspace scope, loopback binding for Core and Bridge, mode 0600 for the Core token, channel-secret handling, and explicit Memory sharing; Kimi also does not auto-approve pending tool requests. Those details support moderate least-privilege and sensitive-data scores. Codex defaults to approvalPolicy=never, so user confirmation is weak. The materials describe major data and channel paths but do not provide a complete data inventory or per-flow account. Dependencies have version ranges and an override list, but no audit or vulnerability-response evidence. Agents can read, write, and execute code and connect to IM, external Core, and scheduling; controls over those external effects are only partly documented. Updates preserve user data and allow a pinned release, but a complete recovery procedure is absent. Contributor attribution and an MIT license are present, while publisher identity and maintenance responsibility remain unclear.

Reliability8 / 14 · 2.9/5

The README, installation docs, CLI, and engine-selection code are broadly consistent about Codex defaults, Kimi/Claude compatibility, ports, and main commands; several capability boundaries are stated. Dependencies can be installed from npm and GitHub, but the manifest includes version ranges and a URL dependency, with no evidence here of supply-chain auditing or offline availability. Bounded install diagnostics, health/doctor/status commands, logs, and troubleshooting steps make common failures reasonably diagnosable.

Adaptability12 / 18 · 3.3/5

The materials target self-hosted personal use and cover Feishu, Telegram, Slack, WeChat, Web, multiple engines, and workspace configuration. Engine differences, the incomplete Windows Core/Web lifecycle, and the unavailable Codex app-server path are stated; the sampled Skills also define their command purposes. Skill triggers are clear, but product-level scenarios, trigger policies, and environmental constraints are not comprehensive enough for full marks.

Convention13 / 18 · 3.6/5

Documentation is organized across setup, configuration, features, reference, and troubleshooting, with many README links. Installation, updates, platform differences, and troubleshooting have concrete steps. CLI and Skill naming is consolidated around metabot, though multiple engines and compatibility paths remain, so complete naming stability is not established. Examples are plentiful, but FAQ evidence is limited; some limitations are disclosed without a comprehensive limitations list. The license file clearly states MIT. The package is versioned 1.3.4, but the sample has no changelog or clear version-governance account. AGENTS.md describes project guidance and the license names MetaBot contributors, but maintenance ownership and update responsibility are only partly established.

Effectiveness7 / 13 · 2.7/5

The product combines a CLI, cross-channel Agent sessions, shared memory, Teams, scheduling, and a local console, which gives it a clear integration use. Command examples, run identifiers, and troubleshooting improve practical output usability. The setup guide explains common costs, but operation requires local services, engine login, channel credentials, and maintenance of several integrations; there is no quantified cost or fit assessment, so value and cost-benefit do not merit full marks.

Verifiability5 / 8 · 3.1/5

Several README claims have corresponding support in installation/security docs, CLI and engine code, or Skill descriptions. CI configuration and sampled tests show some intended engineering verification, but this static review does not infer that tests ran or passed. Documentation often distinguishes supported features, limitations, and configuration guidance, while release and security claims remain primarily project assertions with limited independent or cross-source evidence.

Risks and how to mitigate them
  • Codex defaults to approvalPolicy=never; confirm workspace boundaries, channel visibility, and host isolation before unattended use.
  • The installation materials describe checksum verification; the supplied sample does not show release-signature verification implementation or dependency-audit evidence.
  • Remote Core/Bridge access, IM bots, and scheduling expand data and execution impact; keep services loopback-bound or behind a user-managed authenticated network boundary, and share Memory selectively.
Evidence confidence: Low Reviewed Nov 08, 2026 Reviewed revision f884e605c5fc
Review evidence README.mdAGENTS.mdCLAUDE.mddoubao-tts/SKILL.mdfrontend-design/SKILL.mdmetabot-team/SKILL.mdmetabot/SKILL.mdconcepts/security.mdconcepts/security.zh.mdgetting-started/installation.mdgetting-started/installation.zh.mddocs/troubleshooting.mdsrc/index.tsengines/index.tssrc/index.tsLICENSEpackage.jsonworkflows/ci.ymlworkflows/docs.yml
See the full review method →
Maintain this project? Add the FARS badge to your README

The badge shows the current review score and verdict and links back to this page. It refreshes about once an hour after a re-review.

FARS

[![FARS review](https://followagents.com/badge/metabot.svg)](https://followagents.com/en/agents/metabot)
<a href="https://followagents.com/en/agents/metabot"><img src="https://followagents.com/badge/metabot.svg" alt="FARS review"></a>

Recent releases

Latest entries from the project's GitHub Releases. Notes are quoted from upstream as written, not rewritten or tested by FollowAgents.

  1. v1.3.4
    Release notes excerpt
    What's Changed * 新增 Codex 和 Claude Code 执行中插话：支持 /append 与 /steer by @wengxiaoxiong in https://github.com/xvirobotics/metabot/pull/427 * 新增 Codex 和 Claude Code 手动压缩：保留会话并排队处理消息 by @wengxiaoxiong in https://github.com/xvirobotics/metabot/pull/428 * feat: sync public-safe session Bus, Console, automation and runtime updates by @floodsung in https://github.com/xvirobotics/metabot/pull/431 * release: prepare v1.3.4 patch release by @floodsung in https://github.com/xvirobotics/metabot/pull/432 Full Changelog: https://github.com/xvirobotics/metabot/compare/v1.3.3...v1.3.4
    View on GitHub ↗
  2. v1.3.3
    Release notes excerpt
    What's Changed * feat: add /new reset and /? help aliases by @wengxiaoxiong in https://github.com/xvirobotics/metabot/pull/421 * 修复飞书单独 @机器人无响应：先读取聊天历史再回复 by @wengxiaoxiong in https://github.com/xvirobotics/metabot/pull/426 * fix(agent-teams): keep in-flight turns from failing when the store closes during shutdown by @Ivancheng7 in https://github.com/xvirobotics/metabot/pull/412 * feat(observability): make prompt/response persistence in activity and audit logs opt-out by @Ivancheng7 in https://github.com/xvirobotics/metabot/pull/413 …
    View on GitHub ↗
  3. v1.3.2
    Release notes excerpt
    What's Changed * test: redirect USERPROFILE alongside HOME for Windows by @Ivancheng7 in https://github.com/xvirobotics/metabot/pull/401 * fix: recover undeliverable AskUserQuestion answers by @Ivancheng7 in https://github.com/xvirobotics/metabot/pull/396 * fix: prefer native claude executables on Windows by @Ivancheng7 in https://github.com/xvirobotics/metabot/pull/397 * fix: fall back to metabot-core bootstrap token by @Ivancheng7 in https://github.com/xvirobotics/metabot/pull/398 …
    View on GitHub ↗

FAQ

Must I use Codex?
No. A bot can use Codex, Kimi Code, or a Claude Code-compatible path. Omitting engine defaults to codex.
Do I need a public IP address?
No. The documented Feishu/Lark integration uses a persistent WebSocket connection, while Telegram and WeChat use long polling.
How is the local console accessed?
Open http://localhost:9200 and paste the installer-generated Bearer token stored in ~/.metabot-core/token.
What permissions do agents receive?
They can read, write, and execute code in their configured working directories, so only trusted workspaces and chat participants should be authorized.
Was this page helpful?
View on GitHub ↗ Install ↓

Compare agents like this one

The same FARS review applied across the shortlist this agent qualifies for.

Open Source Coding Agents, Compared

Open source coding agents ranked by FARS review — score, trust, license, language and maintenance side by side, with what each one is and isn't good at.

Claude Code Agents, Compared

Claude Code agents ranked by FARS review — compatibility is read from each agent's own AGENTS.md and documentation, not assumed, and only full support is listed.