Add voice input and image support to opencode-router #625
Description
Summary
Add voice input and image support to opencode-router (Telegram and Slack).
Context
OpenWork is mobile-first and relies on messaging surfaces like Telegram and Slack. opencode-router connects those surfaces to an OpenWork and OpenCode instance: https://github.com/different-ai/openwork/tree/dev/packages/opencode-router.
Scope
- Ingest Telegram voice notes and Slack audio messages, transcribe to text, and pass to OpenCode.
- Ingest images and forward to OpenCode as attachments or uploaded assets.
- Update docs and sample config for these new inputs.
Acceptance
- Voice message to transcript to message in an OpenWork session.
- Image forwarded and accessible in the OpenWork session.
- Works on both Telegram and Slack with clear error handling and logs.