Use this skill whenever the user wants to call StepFun Chat Completions with model step-audio-r1.1 for a non-streaming speech reasoning turn that can take text with optional local audio input and save the returned audio, transcript, and raw response object. Triggers include mentions of StepFun speech reasoning, step-audio-r1.1, non-streaming spoken reasoning through the Chat API, text+audio input to StepFun speech reasoning, or requests to inspect/save the raw response payload from StepFun step-audio-r1.1 runs.
Use this skill whenever the user wants to convert text into speech, generate playable audio from written text, read a sentence aloud, narrate content, create voiceovers, or clone / customize voices, and the backend must be 阶跃星辰 StepFun TTS. Triggers include mentions of 'TTS', 'text to speech', 'step-tts', 'stepfun', '阶跃', '语音合成', '生成语音', '转成语音', '读出来', '念出来', '朗读', '配音', '旁白', '播报', or requests to turn written content into spoken audio using StepFun models. Also use when the user wants to control StepFun voice IDs, speed, volume, emotion/style tags (情绪/风格标签), or call the StepFun voice-clone API with an existing file_id. Do NOT use this skill for speech recognition or audio-to-text tasks; those belong to ASR skills. Do NOT use this skill for Noiz or Kokoro backends.
Use this skill whenever the user wants to convert audio into text, transcribe recordings, recognize spoken content, generate transcripts, extract subtitles from audio, or perform speech-to-text with 阶跃星辰 StepFun ASR. Triggers include mentions of 'ASR', 'speech to text', 'audio to text', 'transcribe', 'transcription', '语音转文字', '音频转文字', '转写', '录音转文字', '听写', '识别语音', '识别音频', or requests to turn spoken audio into written text using StepFun models. Also use when the user wants streaming transcription output, terminology correction prompts, usage statistics, or explicit format settings for PCM/WAV/MP3/OGG input. Do NOT use this skill for text-to-speech, reading text aloud, narration, dubbing, or other text → audio tasks; those belong to TTS skills such as step-tts.