AgentConnect

The open-source, multi-agent alternative to Claude Tag.
@ any agent, wherever work happens.

_{FROM                          WITH                     ANY ACP AGENT}

⭐ Star on GitHub · Documentation · Website · Blog

Use cases · Get started · Community · Why AgentConnect? · Build your stack · Architecture · Development · Explore

Multiple agents, one team—in the tools your team already uses.

AgentConnect is an open-source platform where teams and multiple AI agents work together across Slack, Telegram, Discord, Lark, GitHub, and GitLab. Bring Claude Code, Codex, Grok Build, DeepSeek, Pi, or any ACP-compatible agent into the conversations and workflows your team already has open.

Give each agent a role, then let people and agents collaborate in shared conversations. Agents can call one another, and work can begin from a message, issue, pull request, webhook, or schedule.

Each agent can have its own model, workspace, memory, MCP servers, skills, repository access, and sandbox policy. Fine-grained controls separate who may use an agent from who may see its sessions, while one console keeps the team's agents, integrations, and permitted work in view.

What teams do with AgentConnect

• Triage issues together. People and agents investigate in one shared thread, bring in the right specialists, and keep the fix and verification visible from start to finish.
• Customized code review. Run a general reviewer on every pull request, then bring in architecture or security reviewers only when a change needs them. Each reviewer can use its own model, instructions, repository access, tools, and sandbox policy.
• Support across trusted workspaces. Start a support conversation in Telegram, involve engineering from a trusted Slack workspace, and return the resolution where the conversation began.
• Run recurring operations. Start work from a schedule or webhook, bring exceptions into a shared conversation, and keep the human decision visible.
• Keep private forks current. Subscribe to upstream changes through GitHub, a GitHub subscription in Slack, webhooks, or schedules. Let agents assess the impact, prepare and test relevant updates, and bring them to the team for review.

Get started

Pick the path that matches what you want to do.

Self-host locally

Start the Web console, Control Plane, Relay, and PostgreSQL with Docker Compose:

git clone https://github.com/agentconnect-md/agentconnect.git
cd agentconnect
docker compose up -d --pull always

Open http://localhost:3000, add a daemon from the console, run its generated command, and create your first agent. The default stack listens only on 127.0.0.1 and uses local no-auth mode for evaluation.

For Kubernetes, install the official Helm chart in charts/agentconnect, published to oci://ghcr.io/agentconnect-md/charts/agentconnect on every release. For authentication, public URLs, Linux sandbox requirements, provider apps, image pinning, secrets, and optional Mem0 configuration, follow the AgentConnect OSS guide.

Community

If AgentConnect looks useful, consider giving the repository a ⭐. It helps more teams discover the project.

Questions or ideas? Join the Slack community, open a GitHub issue, or follow project updates on X and YouTube.

Why AgentConnect?

AI agents are taking on work across the team, but most still live in individual terminals. AgentConnect brings them into the team's shared workflows:

• Work as one team. Create agents with different roles and let them call on one another, while people follow along in the conversations where the work happens.
• Keep work where it happens. Connect agents to Slack, Telegram, Discord, and Lark, or to repositories and workflows on GitHub or GitLab—even across trusted messaging workspaces.
• Configure each role independently. Give every agent the model, workspace, memory, MCP servers, skills, repository access, and sandbox policy its work requires.
• Control access for people and agents. Manage agent access separately from session visibility, link supported social identities, and decide which repositories, tools, and other agents each agent may reach.
• Stay provider-neutral. Run Claude Code, Codex, Grok Build, DeepSeek, Pi, and other ACP-compatible agents side by side.
• Stay in control. Self-host the Apache-2.0 stack, run agents in your environment, and keep execution and workspaces under your control.

Build your stack

Layer	Options
Agent runtimes	Claude Code, Codex, Grok Build, DeepSeek, Pi, and other ACP-compatible runtimes
Channels	Slack, Telegram, Discord, Lark / Feishu, and webchat
Triggers	GitHub and GitLab events, generic webhooks, and schedules
Memory	AgentConnect-managed memory, supported runtime-native memory, external providers, or Off
Tools and apps	Custom MCP providers and OpenConnector-backed services
Knowledge and skills	Reviewed organization knowledge, immutable managed skills, and Git-based skill sources with per-agent enablement
Team controls	Agent access, session visibility, linked social identities, repository and tool access, and agent visibility
Agent configuration	Independent runtime, model, workspace, credentials, sandbox policy, and placement per agent

Architecture

Component	Responsibility
Daemon	Runs placed agents over local ACP, owns workspaces and session state, maintains direct platform connections and schedules, and sends model-provider traffic directly
Relay (optional)	Accepts callback-based ingress and webchat, proxies centrally managed MCP and OpenConnector access, and forwards message ingress directly to the owning daemon without durable storage
Control Plane + Web UI	Manages authentication, configuration, placement, permissions, metadata, and observability; stores explicitly approved organization knowledge/skill revisions and otherwise proxies bounded daemon reads on demand

Live platform messages and ACP update streams stay on the daemon/relay data plane. Apart from explicitly approved organization knowledge and bounded skill bundles, the Control Plane stores coordination metadata—not message bodies, attachment bytes, pending Dream proposals, or ACP session streams. If it is temporarily unavailable, established sessions and daemon-local schedules continue; new assignments and configuration changes resume after reconnection.

See the daemon-centric architecture for the complete message paths, trust boundaries, and failure model.

Development

Development requires Node >= 24.12.0 and pnpm 11. Docker is required for the Control Plane integration tests.

pnpm install
pnpm dev          # run all packages in parallel
pnpm build        # build all packages
pnpm typecheck    # type-check all packages
pnpm lint         # lint the workspace
pnpm format:check # check formatting
pnpm test         # test all packages

# single package
pnpm --filter @agentconnect.md/daemon dev
pnpm --filter @agentconnect.md/control-plane dev
pnpm --filter @agentconnect.md/web dev

For a complete local Control Plane and PostgreSQL development setup, follow the Control Plane quickstart.

Monorepo layout

This repository is a pnpm workspace. Product packages live under packages/:

Package	Path	Role
@agentconnect.md/cli	packages/cli	Stable agentconnect entry point, daemon lifecycle, and upgrades
@agentconnect.md/connection	packages/connection	Shared WebSocket transport, correlation, backoff, and keepalive
@agentconnect.md/control-plane	packages/control-plane	Orchestration, registry, authentication, and Web UI BFF
@agentconnect.md/daemon	packages/daemon	Edge message processing and agent execution unit
@agentconnect.md/memory-plugin-mem0	packages/memory-plugin-mem0	Mem0 Cloud and OSS memory-plugin profiles
@agentconnect.md/message	packages/message	Pure platform message normalization
@agentconnect.md/protocol	packages/protocol	Shared daemon, relay, and Control Plane wire contracts
@agentconnect.md/relay	packages/relay	Callback ingress, webchat, and centralized MCP proxy
@agentconnect.md/setup	packages/setup	Browser-based self-hosting and provider App administration
@agentconnect.md/web	packages/web	Next.js configuration and monitoring console

Explore

• Public documentation
• Self-host AgentConnect OSS
• Architecture and detailed designs
• CLI and daemon lifecycle
• Setup Server
• Daemon configuration
• Config-file secrets
• Product conventions
• Add-on evaluation harness

License

AgentConnect is available under the Apache License 2.0.