Skip to content

chore: Add tool usage + session audio recording#36

Merged
Ruthvik-1411 merged 11 commits intomainfrom
gemini_live_custom
Jul 4, 2025
Merged

chore: Add tool usage + session audio recording#36
Ruthvik-1411 merged 11 commits intomainfrom
gemini_live_custom

Conversation

@Ruthvik-1411
Copy link
Owner

Changelog:

  1. Handling tool calls is added for gemini live events.
  2. A decorator was built to convert any function tool to openapi specified declaration schema. This abstracts the developer from creating the entire function description again. They can just write the function very cleanly.
  3. Minor changes in the UI to display tool call events in the transcript.
  4. Added support for session audio recordings. Now, all the user and audio speech is recorded and saved after the session ends.

Todo:

  1. Abstract the gemini handler class.
  2. Implement the actual call_function in gemini handler, so handler handles the function calling.
  3. Create a simple definition for agent and tools like adk root_agent and tools.
  4. Add logic for combining user audio and model audio into a single mp3 file.

@Ruthvik-1411 Ruthvik-1411 merged commit 06f4a42 into main Jul 4, 2025
2 checks passed
@Ruthvik-1411 Ruthvik-1411 deleted the gemini_live_custom branch July 4, 2025 18:46
@Ruthvik-1411 Ruthvik-1411 restored the gemini_live_custom branch July 4, 2025 18:46
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant