This is different from the standard setup in the Quickstart, where you connect your own Teams bot and callers dial it. On the community tier there is no Azure bot to create: a shared StandIn bot joins a public meeting you generate.
What you need
- A running OpenClaw or Hermes agent, reachable over the internet at its voice endpoint
wss://<your-host>/voice/msteams/stream, plus the shared secret it was configured with. - A sign-in at standin.komaa.com. No Azure bot and no Teams admin setup are required.
Steps
Connect your agent
Sign in at standin.komaa.com/sandbox and choose Connect your agent. Enter your agent’s
wss:// voice URL and its shared secret, then click Connect. StandIn creates a private room with its own *.standin.komaa.com address.Generate a meeting link
In the room, click Generate meeting link. StandIn creates a Microsoft Teams meeting and returns a shareable
https://teams.microsoft.com/... join link.Share the link
Share the join link anywhere (a chat, a page, an email). Joiners need nothing installed, because the meeting is public.
Talk to the AI
When someone opens the link and joins the meeting, the StandIn bot joins automatically and bridges the audio to your agent. Speak, and your AI answers by voice, with an animated avatar (Jarvis) on screen. When everyone leaves the meeting, the bot leaves too, and it rejoins the next time someone joins.
Limits
The community tier is metered so the shared infrastructure stays fair for everyone:- 5 minutes of bot time per day, per room. Each room has its own daily budget; when it runs out, the bot leaves and returns the next day.
- Up to 5 new rooms per day per account. Reuse your existing room any time (it stays live for its 7 days at no extra cost); only creating a brand-new room counts toward this daily limit.
- One bot per meeting (a second join trigger is ignored).
- The room link is valid for 7 days; after that you generate a new one.
How your AI connects
Every meeting routes to the agent you connected. StandIn matches the meeting to your room and bridges that call’s audio to yourwss:// endpoint over a per-call authenticated connection. Your agent does the speech, language, and voice; StandIn only carries the Teams media. Keep your agent reachable at its wss:// URL, or the bot has nothing to connect to.