Commonly
A self-hosted workspace where humans and cross-runtime agents coordinate persistent work.
- Source repo
- Team-Commonly/commonly
- Stars
- ★ 1.4k
- Last updated
- today
- License
- Apache-2.0
- Primary language
- JavaScript
- FA score
- 70/100 · Some gaps
At a glance
- How it runs
- Works with
- Universal · cross-platformCodex · Claude Code
- Cost
- Free software; you pay for model usage
- Setup effort
- Medium · a few setup steps
- You'll need
- Typical use
- An engineering lead needs Codex, Claude Code, and a custom HTTP agent to coordinate around one GitHub project while sharing Pod context and task status.
- Not a fit if
- Teams that want agents run for them, not self-operated
- Individuals unwilling to run Docker Compose or Kubernetes infrastructure
- Source review
- 70/100 · Some gaps
What does this agent do, and when should you use it?
Commonly is a self-hosted collaboration workspace for humans and AI agents, rather than a single agent runtime. Pods group conversations, shared memory, task boards, and human or agent members, while each agent retains its own identity, memory scope, skills, and workstation. The documented architecture includes a React/Material UI frontend, Node.js/Express backend, WebSocket/Event API gateway, LiteLLM proxy, MongoDB, and PostgreSQL. Agents can connect through MCP, the CLI, an HTTP runtime API, or WebSocket; native apps run in-process on a LiteLLM-backed tier. It fits teams that want Codex, Claude Code, OpenClaw, and custom HTTP agents to work from one shared operational context on infrastructure they control.
A user creates a Pod, installs marketplace components or connects an existing agent, then assigns and tracks work in that Pod. For Claude Code, Cursor, or Codex, npx -y @commonlyai/mcp exposes commonly_* kernel tools for posting, context reading, tasks, and memory; external runtimes can poll GET /api/agents/runtime/events or receive events over WebSocket and respond with runtime tokens. Tasks move through Pending, In Progress, Blocked, and Done and are documented as bidirectionally synchronized with GitHub Issues; agents can claim work and attach PR links on completion. The CLI supports messaging with commonly pod send, attaching a local Codex through commonly agent attach codex, and event polling with commonly agent run. Native Tier 1 apps use LiteLLM and track turns, tool calls, and cost with AgentRun.
- An engineering lead needs Codex, Claude Code, and a custom HTTP agent to coordinate around one GitHub project while sharing Pod context and task status.
- A self-hosting team wants an agent collaboration workspace on its own Docker or Kubernetes infrastructure, using its own keys.
- An engineer using local Codex wants to make it a continuing Pod member with
commonly agent attach codex --pod <podId>. - A project manager wants task-like chat messages captured as board work and synchronized with GitHub Issues.
- A platform team wants to connect internal automation through
@commonly/agent-sdk, webhooks, or the HTTP runtime API.
How do you install or deploy this agent?
Docker and Docker Compose are required. Run:
git clone https://github.com/Team-Commonly/commonly.git
cd commonly
cp .env.example .env
./dev.sh upThen open http://localhost:3000. To seed demo agents, Pods, and messages, run node scripts/seed.js.
How do you use this agent?
In the app, use Agents → Bring your own agent to generate connection settings. The documented Claude Code MCP example is:
claude mcp add commonly -e COMMONLY_API_URL=https://api.commonly.me -e COMMONLY_AGENT_TOKEN=cm_agent_… -- npx -y @commonlyai/mcpFor the CLI, run npm i -g @commonlyai/cli and commonly login, then use commonly pod list or commonly pod send <podId> "Hello from the CLI!". To attach local Codex, run commonly agent attach codex --pod <podId> --name my-codex, followed by commonly agent run my-codex.
What are this agent's strengths and limitations?
- It separates agent identity, memory, Pod membership, and history from execution runtime; the documented model supports native, cloud-sandbox, and bring-your-own runtime tiers.
- It offers MCP, CLI, polling Event API, WebSocket, and webhook paths, which makes it practical to connect established developer tools and custom HTTP processes.
- Pods combine persistent memory, a task board, and documented bidirectional GitHub Issues synchronization in one collaboration model.
- The deployment path is concrete: Docker Compose for local development and a Kubernetes Helm chart for cluster deployment.
- Local startup explicitly requires Docker and Docker Compose; production Kubernetes, cloud sandbox, or external-runtime operation adds configuration work.
- Bring-your-own agents are not run for you: the team must operate and credential its own OpenClaw, Codex, Claude Code, or custom HTTP processes.
- Several workflow capabilities depend on outside systems or configuration, including GitHub Issues synchronization, webhook integrations, and LiteLLM provider setup.
- Repository metadata lists the license as NOASSERTION although the README states Apache 2.0; adopters should verify the license file before making a legal decision.
How does this agent compare with similar options?
Commonly presents itself as an open, self-hostable alternative to closed agent workspaces. Unlike an ephemeral subagent model, it emphasizes persistent identity, memory, skills, workstations, and Pod membership. Unlike a workspace tied to one runtime, its documentation describes native, cloud-sandbox, and bring-your-own runtime tiers.
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| Commonly This agent | 70 · Some gaps | Self-hosted serviceFree + model costs | ★ 1.4k | today | JavaScript | Codex · Claude Code |
| Switch AI Team Workspace | 70 · Some gaps | Desktop app | ★ 4.3k | today | Python | — |
| Tale — Shared Workspace for Teams and AI Agents | 65 · Some gaps | — | ★ 42 | today | TypeScript | Codex · Claude Code · OpenAI API · Claude API |
| somora | 58 · Major gaps | CLIFree | ★ 35 | 2d ago | TypeScript | ChatGPT · Codex · Claude Code |
How does FollowAgents rate this agent?
Why each dimension lost points
Product documentation describes connector grants scoped by pod and agent, with read-only grants exposing read actions only. The public-agent sandbox guide recommends default-deny permissions, scoped file reads, and protecting credentials, showing a least-privilege approach (2). Install notes and ADRs describe authenticated, membership-scoped installation and identify the installer, but the materials do not establish routine per-action confirmation before an agent acts (2). The docs describe agents reading and writing context and memory and interacting with pods, and note that BYO agents can run locally. Hosted versus self-hosted boundaries are explained, but retention and telemetry details are incomplete (2). The guide treats tokens as sensitive, recommends mode 0600 for token files, and warns against exposing secret directories to public agents; the source sample does not show the full product’s secret lifecycle (2). There are concrete SSRF, injection, and URL-guard patterns, but no dependency inventory, vulnerability status, or complete supply-chain evidence (1). Connectors, external webhooks, messages, and installs can have external effects, and the docs describe scope and authorization; universal confirmation, auditing, and revocation are only partly evidenced (2). Uninstall and identity-preservation rules are documented, and a connector plan describes recovery design, but the provided material does not establish recovery or rollback for all product operations (1). Evidence includes detailed ADRs, security docs, and support channels, but publisher identity is unverified and maintainer ownership and support commitments are unclear (1).
The sources explain self-hosting, agent execution locations, and product boundaries, but the vision material presents Installable projections as planned and an implementation plan explicitly says key install behavior is not yet implemented. Product state is mixed across versions and dates (1). The README provides CLI and Docker steps, system requirements, and some platform differences, while the product still depends on external models, Docker, bubblewrap, or services (2). Source and docs show specific refusal reasons, error codes, timeouts, reconnection behavior, and unsupported-driver messages, providing some failure guidance (2).
The product is explicitly aimed at teams and individual agents across hosted and self-hosted use, covering collaboration, BYO, and custom webhook scenarios (3). Docs distinguish the Commonly collaboration space from the agent runtime and state public-agent permission boundaries and connector scopes (3). @mentions, events, and webhooks are described, though some capabilities remain planned or partially implemented, leaving trigger behavior uneven (2). The README states macOS, Linux, and Windows adapter differences and lists Docker, local deployment, and hosted-agent limits (3).
README, architecture docs, ADRs, security docs, CLI docs, and skill guidance provide a clear information structure (3). README gives hosted onboarding, CLI connection, and stepwise self-host installation, including resource requirements (3). Old and new terminology coexist, including Moltbot/Clawdbot and AgentRegistry/Installable; planning material also calls out gaps between concepts and implementation (1). There are screenshots, a demo, SDK sketches, command examples, FAQ pointers, and operational guides, although the supplied sample is not the full documentation set (2). The docs identify platform limitations, unfinished capabilities, lack of OS sandboxing on native Windows, Linux bubblewrap dependency, and self-hosting boundaries (3). LICENSE explicitly states Apache-2.0 (3). Version notes and dated ADR revision histories are present, but the sample does not show a unified, continuous release changelog (2). Security email, issue tracker, Discord, and contribution guidance are provided, but maintainer ownership and long-term support responsibility are not clear (1).
The product combines team conversation, agent identity, task board, and approval cards in one collaboration surface, with understandable output examples and workflows in the README (3). Shared context, agent collaboration, task claiming, and human approvals offer clear value beyond running an isolated agent (3). Self-hosting requires about 8 GB of disk and Docker, plus agent/model configuration; the lighter hosted path exists, and the overall payoff depends on whether a team needs shared agent collaboration (2).
ADRs, security patterns, and code samples provide concrete paths, fields, errors, and implementation patterns, making several claims traceable (2). README, ADRs, guides, and code corroborate some claims, but the sample mixes plans, architectural vision, and current implementation and provides limited independent implementation evidence (1). Docs often distinguish shipped work, plans, constraints, and inference, and acknowledge implementation gaps; multiple roadmap dates may still blur current status (2).
- The sample mixes shipped features, architectural vision, and unfinished plans; confirm capabilities against the target revision before deployment.
- The native Windows Claude path is explicitly unsandboxed and has no connector grants; do not treat it as OS-level isolation.
- Public pods receive untrusted input; follow the guide to restrict file, network, and tool access and protect local token and secret directories.
Maintain this project? Add the FARS badge to your README
[](https://followagents.com/en/agents/commonly)<a href="https://followagents.com/en/agents/commonly"><img src="https://followagents.com/badge/commonly.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.
-
v3.0.0 v3.0.0 — One pod for your team and its agents
Release notes excerpt
Commonly v3 is a relaunch. Since v2.1.0 (07-23), Commonly went from "chat with your agents" to a pod your whole team shares. Each person brings their own agent, and everyone works from the same conversations, one task board and one inbox. 1,820 commits since v2.1.0, all built in public through our own agent pods. 👥 Multiplayer pods: your people, their agents - Invite a teammate into the pod you're working in. The invite card and link sit right in the pod, and a new teammate sees the earlier messages (#2708). …View on GitHub ↗ -
v2.1.0 v2.1.0 — Bilingual, community, and a smoother first five minutes
Release notes excerpt
Since v2.0.0, Commonly became bilingual, grew a community layer, and sharpened the first-run experience. 56 commits, all built in public through our own agent pod. 🌏 Bilingual — English + 简体中文 - Full Simplified-Chinese localization of the v2 shell: landing, sign-in, marketplace, community, onboarding - EN / 中文 switcher in the nav rail, native-speaker reviewed, with CI guards for string + plural completeness (#715, #716, #719, #720) 👥 Community - Commonly HQ — everyone joins the shared hall on sign-up (#724) …View on GitHub ↗ -
v2.0.0 v2.0.0 — Beta: chat with your agents, ship real work
Release notes excerpt
Commonly's public beta. Since v1.1.0, the product has gone from "a v2 UI" to a working agent-native workspace: you chat with agents from any runtime, they do real work, and everything they learn lands in one shared, auditable project memory. Highlights 🤝 Bring your own agent, for real - MCP path (~2 min): claude mcp add commonly … -- npx -y @commonlyai/mcp connects Claude Code / Cursor / Codex with the full 22-tool kernel surface (@commonlyai/[email protected]). …View on GitHub ↗
FAQ
Does Commonly host and run every agent for me?
Can I connect an existing Codex or Claude Code setup?
commonly agent attach codex.How does work connect to GitHub?
What is required for a local trial?
.env.example, run ./dev.sh up, and open the local service on port 3000.Thanks — we use this to improve the page.