Step Asr

Transcribe audio files to text via Step ASR streaming API (HTTP SSE). Supports Chinese and English, multiple audio formats (PCM, WAV, MP3, OGG/OPUS), real-ti...

This skill appears to implement exactly what it claims: it base64-encodes a local audio file and sends it to the Step/StepFun ASR endpoint using the STEPFUN_API_KEY. Before installing, consider: (1) Privacy—audio you transcribe is uploaded to StepFun, so avoid sending sensitive audio unless you trust the service and key; (2) API key handling—store the STEPFUN_API_KEY securely and avoid hardcoding it; (3) Resource use—the script reads the entire file into memory (may be large for long audio); (4) Review network access—the script contacts api.stepfun.com, which matches the description; (5) If you need different behavior (streaming chunks from disk rather than single large upload), inspect/modify the script. Overall the skill is coherent with its stated purpose.

Type: OpenClaw Skill Name: step-asr Version: 1.0.0 The skill bundle is designed to transcribe audio files using the Step ASR API. The `SKILL.md` provides clear instructions and usage examples, and the `scripts/transcribe.py` script implements this functionality by reading a specified audio file, base64-encoding it, and sending it along with an API key (read from `STEPFUN_API_KEY` environment variable) to the `https://api.stepfun.com/v1/audio/asr/sse` endpoint. There is no evidence of unauthorized data exfiltration, malicious execution, persistence mechanisms, prompt injection attempts against the OpenClaw agent, or obfuscation. The file writing capability (`--out` argument) is for saving transcription results, which is a legitimate function of the skill, and does not indicate malicious intent within the skill itself.