Input
ExecutionTrigger
Data/Microsoft/Copilot Studio
Send a message to a published Copilot Studio agent and stream its reply. Leave the conversation ID empty to start a new conversation; pass the returned ID to continue it. Billed to the agent's tenant in Copilot Credits.
Scores range from 0 to 10. Higher values mean more impact, exposure, or operational weight.
Trigger
Copilot Studio provider with a delegated Power Platform token
Copilot Studio agent reference (from the Copilot Studio Agent node)
`https://{environment host}/copilotstudio/…/bots/{schema name}/conversations`, without query.
The user's message
Conversation to continue; leave empty to start a new one
Language of the conversation, e.g. en-US or de-DE
Fires for every streamed text or progress chunk
Streamed chunk: answer text as content, progress updates as reasoning
The agent finished its turn
The call failed
The agent's reply with citations as annotations
Ordered structured content returned by Rig when the assistant response contains media. Text-only responses continue to use `content` alone for OpenAI compatibility.
Pass this into the next call to continue the conversation
Quick replies the agent offered; send one back as the next message
Adaptive Cards and other attachments the agent sent
Citations from generative answers, as chat attachments
The agent asked a question and waits for the user's answer
The agent ended the conversation; start a new one for the next message
Every raw Bot Framework activity received in this call
Why the call failed