Converts MP3 audio into Feishu's voice message format and sends it as a native voice clip (not a downloadable file) to a user or chat. Typical workflow: verify environment, convert MP3→AMR (ffmpeg), upload to Feishu to obtain a file_key, then send with msg_type: "voice" and confirm delivery. Useful for broadcast voice notifications, private voice messages, and automating voice-based alerts.
Inferred: agents with Bash/Python ability and network access (Claude Code, Copilot/Code agents, generic Python-capable assistants).
Feishu Voice Sender converts MP3s to AMR and sends them as native Feishu voice messages. The SKILL.md is well-written with clear triggers, guardrails, and bilingual docs (Chinese/English), but the actual Python script (feishu_voice_sender.py) is not bundled — only referenced. No scripts were available to run or audit. The skill describes a real Feishu API workflow but is incomplete without the implementation.
No bundled scripts to audit — the skill is a documentation-only SKILL.md pointing to an external Python script. The referenced script path (skills/feishu_voice_sender/feishu_voice_sender.py) is not included in the repo. Security score is high since there's nothing executable to flag. Quality and architecture scores are moderate because the instructions are good but the implementation is missing.