senseaudio-open-platform

# SenseAudio Open Platform Use this skill for implementation work around `api.senseaudio.cn` and `wss://api.senseaudio.cn`. ## Quick Routing - TTS text-to-speech: read `references/tts.md` - ASR speech recognition and audio analysis: read `references/asr.md` - Realtime agent session lifecycle: read `references/agent.md` - Video generation and storyboard APIs: read `references/video.md` - Voice plan levels and clone constraints: read `references/voice.md` Load only the relevant reference file(s) for the user task. ## Default Workflow 1. Confirm capability and protocol: - TTS: HTTP sync, SSE stream, or WebSocket stream. - ASR: HTTP file transcription or WebSocket realtime stream. - Agent: REST lifecycle plus external realtime media channel. - Video: upload/create/poll/storyboard pipeline. 2. Build a minimal valid request first: - Add `Authorization: Bearer <API_KEY>`. - Set required fields only. - Validate model and endpoint compatibility. 3. Add advanced options only when asked: - Voice tuning, dictionary, translation, diarization, timestamps, storyboard edits. 4. Parse responses safely: - Check status fields before using payload data. - For TTS, decode hex audio to bytes before saving or playback. - For streams, aggregate chunks and finalize on terminal event/status. 5. Add production hardening: - Timeout and retry strategy for transient failures. - Explicit handling of auth, quota, parameter, and not-found errors. - Structured logs with `trace_id` or session identifiers when available. ## Implementation Rules - Keep API keys in environment variables, never hardcode secrets. - Prefer curl first for reproducibility, then provide language SDK code. - For WebSocket flows, enforce event ordering from the reference docs. - Keep examples copy-paste runnable. - If user provides invalid parameter combinations, explain the exact fix. ## Output Checklist When producing implementation output, include: 1. Chosen endpoint and protocol. 2. Minimal request example. 3. One production-ready version (language requested by user). 4. Error handling and response parsing notes. 5. Any model-specific constraints that apply.