- openai_compat: extract reasoning_content from streaming SSE deltas,
benefiting providers like DeepSeek, Mimo, and Kimi that return
chain-of-thought content in streamed chunks
Media handling refactor (addresses review feedback):
- agent: centralize video/audio inline decisions in resolveMediaRefs
with provider capability gating via VideoCapable/AudioCapable interfaces
- agent: add mediaInlinePolicy + mediaInlinePolicyFromProvider() to probe
provider capabilities before encoding media as data URLs
- agent: add size guard via encodeMediaToDataURL (respects MaxMediaSize)
- providers: add VideoCapable and AudioCapable interfaces (same pattern as
ThinkingCapable/NativeSearchCapable)
- openai_compat: implement SupportsVideo/SupportsAudio for Mimo, Qwen,
OpenAI (host and providerName detection)
- httpapi: delegate SupportsVideo/SupportsAudio through HTTPProvider
- httpapi: GeminiProvider declares video + audio support
- common: add video_url serialization in SerializeMessages (now reachable
via the centralized pipeline); fix missing continue after audio branch
- Tests for capability detection, inline policy, oversized media skip,
and provider-unsupported media fallback to path tags
- Add `AudioModelTranscriber` for model-based audio transcription via LLM providers
- Support selecting a transcription model with `voice.model_name` in config
- Keep Groq transcription as a fallback and move it into dedicated files with focused tests
- Serialize `data:audio/...` media as input_audio for OpenAI-compatible providers
- Improve transcription logging by rendering error fields as strings
- Add coverage for transcriber detection, audio-model behavior, provider audio serialization, and Groq transcription
Fixes#1890.
Problem:
During subturn context limit or truncation recoveries, the recovery loops repeatedly
called `runAgentLoop` with the same or modified `UserMessage`. Because `runAgentLoop`
unconditionally adds the `UserMessage` to the session history, this resulted in:
1. Duplicate User Messages polluting the history upon `context_length_exceeded` retries.
2. The possibility of injecting empty User Messages if `opts.UserMessage` was artificially blanked out to work around the duplication.
3. Messy or duplicate entries during `finish_reason="truncated"` recovery injections.
Solution:
- Introduce `SkipAddUserMessage` boolean to `processOptions` to explicitly control whether the agent loop should write the user prompt to history.
- Add an explicit `opts.UserMessage != ""` check in `runAgentLoop` to prevent polluting history with empty message content.
- In `subturn.go`'s recovery loop, set `SkipAddUserMessage: contextRetryCount > 0` to skip writing the user message on context
* Add support for azure openai provider
* Add checks for deployment model name
* Apply suggestion from @Copilot
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Addressing @Copilot suggestion to remove the init() function which seemed redundant
* Fix readme
* Fix linting checks
---------
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
