AgentConnect, a new open-source platform, launched today to enable teams and multiple AI agents to collaborate across the communication and development tools they already use. The platform aims to solve a common problem: AI agents, while increasingly capable, typically work in isolation, with sessions confined to individual terminals and context remaining fragmented. Teams often build separate bots and integrations for each workflow, and products that add a team layer usually require moving work into a dedicated workspace or tie to a single agent provider.
AgentConnect brings agents directly into existing team conversations. Teams can connect Claude Code, Codex, and other ACP-compatible agents, assign each a distinct role, and enable collaboration through shared conversations and agent-to-agent calls. The platform is designed to let humans and agents work together seamlessly.
“AI agents become much more useful when they can work with the rest of the team, not just one person in one terminal,” the AgentConnect team said in a statement. “AgentConnect brings agents with different roles into the conversations where work is already happening.”
The practical applications are immediate. For example, in incident response, a team can ask a triage agent to inspect context, a code specialist to trace a suspected regression, and a verification agent to check proposed results—all while humans add context, redirect work, and follow the diagnosis in the same Slack thread. This model extends beyond incidents: a support issue can move from a customer-facing Telegram conversation into a trusted Slack workspace without manual context copying. Engineering teams can combine general and specialist reviewers for pull requests, with inline GitHub comments and independent check results. Scheduled and webhook-driven work can also continue in a shared conversation when human input is needed.
AgentConnect includes several key features. It supports multiple agents in one team, allowing configuration of distinct models, workspaces, memory, MCP servers, skills, repository access, and sandbox policies for each agent. It works across existing tools, connecting to Slack, Telegram, and Discord, and can start work from conversations, GitHub pull requests, webhooks, or schedules. It offers fine-grained access control, managing agent access separately from session visibility, and allows linking multiple social accounts to one member profile for consistent permissions. The platform is open source under Apache 2.0 and can be deployed with Docker Compose or Kubernetes, ensuring agent execution and workspaces remain in the team's environment.
AgentConnect OSS is available today. The source code can be explored and starred on GitHub, and more information is available at agentconnect.md, with documentation also accessible.
This launch is significant because it addresses a key limitation in AI agent adoption: integration into real-world team workflows. By allowing agents to operate within existing communication tools, AgentConnect could accelerate the use of AI agents in collaborative environments, potentially impacting how teams handle incident response, customer support, and code reviews. The open-source nature of the platform may also foster community-driven improvements and wider adoption.
