The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/send_feishu_voice.sh executes system commands including ffmpeg, curl, and edge-tts. These are used to transcode audio files, interact with the Feishu Open Platform APIs, and generate speech from text respectively.
[DATA_EXFILTRATION]: The script accesses the local file ~/.openclaw/openclaw.json to read Feishu appId and appSecret credentials. This access is required for the skill's primary function of authenticating and sending messages via Feishu.
[EXTERNAL_DOWNLOADS]: The skill instructions and scripts recommend the installation of the edge-tts Python package if it is not already present on the system.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-supplied text to generate audio content, which is a potential injection surface if the text contains tool-specific control characters.
Ingestion points: The text argument passed to the send_feishu_voice.sh script.
Boundary markers: The script uses double-quotes for variable interpolation but lacks explicit instruction boundary markers.
Capability inventory: The script has the ability to execute shell commands (edge-tts, ffmpeg) and perform network requests via curl to Feishu's API.
Sanitization: The script employs shell double-quoting ("$TEXT") to mitigate basic command injection and uses Python's json.dumps for safe JSON construction when sending data to the API.

feishu-voice-tts