AssemblyAI advanced speech transcription

**Judgement**: Strong, agent-focused integration with AssemblyAI that is well-suited for complex transcription workflows, structured exports, and downstream automation. Best used when a user explicitly wants AssemblyAI or needs capabilities beyond generic speech-to-text. **What it does well** - Wraps AssemblyAI transcription, language detection, code switching, and Speech Understanding into a non-interactive Node CLI ideal for OpenClaw-style agents. - Produces stable, machine-readable outputs (Markdown + normalized agent JSON + manifest) that are easy to chain into follow-on tasks. - Supports diarization, speaker role/name mapping (manual + AssemblyAI identification), translation, subtitles (SRT/VTT), and LLM Gateway–based structured extraction. - Provides clear recipes for common scenarios (unknown language, known language with universal-3-pro, meetings with speaker labels, translation, LLM extraction). **Main risks / limitations** - All audio and text are sent to AssemblyAI, which may be unsuitable for highly sensitive data or strict data-residency requirements (mitigated partly by EU endpoint support, but still a third-party processor). - Dependent on AssemblyAI availability, latency, and pricing; failures or quota issues in the external API will directly affect this skill. - Requires a Node environment and correct API key/region configuration; misconfiguration can cause subtle failures or inconsistent behavior. - Rich option set (speaker maps, understanding, LLM Gateway, multiple exports) increases configuration complexity for simple use cases. **Recommended scenarios** - Users explicitly asking for AssemblyAI, or for features like diarization + speaker naming, translation, subtitles, or AssemblyAI’s Speech Understanding. - Agent workflows that need robust, repeatable transcript bundles (Markdown + JSON + manifest) to feed downstream summarization, extraction, or reasoning agents. - Multi-language or uncertain-language audio where automatic routing and language detection are valuable. - Pipelines that already rely on AssemblyAI and want a stable, CLI-based integration with minimal external dependencies.