agentconnect-md/agentconnect ? reverse-engineered prompt

Build me an open source app that lets a team talk to multiple AI agents inside the tools they already use, like Slack, Discord, Telegram, GitHub, GitLab, Gitea, Linear, and webhooks.

I want people to be able to tag an agent in a conversation, give it a role, and have more than one agent help out in the same thread. The agents should be able to respond, call each other if needed, and remember what they learned so the next task has context. It should also handle work that starts from an issue, pull request, message, schedule, or webhook, and keep the whole back and forth visible to the team.

Make it feel like a real product, with a clean docs site or landing page, a simple setup flow, and a working local dev environment with Docker. If you need to, look up the current docs for the integrations and agent protocols online before wiring things up.