The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external URLs which could contain hidden instructions within the audio content. \n
Ingestion points: Data enters the agent context through the fetch_html and download_file functions in scripts/xiaoyuzhou_transcribe.py when processing a user-supplied URL. \n
Boundary markers: The skill does not implement delimiters or instructions to ignore embedded commands for the generated transcript or subtitle output. \n
Capability inventory: The script has the capability to perform network requests using the requests library and write files to the local disk as seen in write_srt and write_txt within scripts/xiaoyuzhou_transcribe.py. \n
Sanitization: No sanitization or filtering is performed on the transcribed text produced by the model. \n- [EXTERNAL_DOWNLOADS]: The download_file function in scripts/xiaoyuzhou_transcribe.py fetches audio files from external URLs derived from the episode page's metadata. \n- [COMMAND_EXECUTION]: The instructions in SKILL.md require the execution of a Python script that performs file system and network operations based on user-provided arguments.

xiaoyuzhou-transcribe