The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands to process media and manage files.\n
Runs mkdir -p x_posts_output to create directories (SKILL.md).\n
Executes ffmpeg with variable parameters (-ss <start_time>, -i "<video_file>", -t <duration>) extracted from the transcription (SKILL.md).\n
Calls internal Python scripts (parse_transcription.py, capture_screenshot.py, fetch_logos.py) passing arguments derived from transcript content (SKILL.md).\n- [EXTERNAL_DOWNLOADS]: The skill facilitates automated network requests to external sources.\n
Uses capture_screenshot.py to visit and screenshot URLs mentioned in the transcript (SKILL.md).\n
Uses fetch_logos.py to download company logos based on domains extracted from the transcript (SKILL.md).\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to how it processes external data.\n
Ingestion points: Reads and parses parsed.json or out.json transcription files which contain untrusted content (SKILL.md).\n
Boundary markers: No explicit delimiters or instructions are used to distinguish between data and potential commands within the transcript (SKILL.md).\n
Capability inventory: Possesses the ability to execute shell commands (ffmpeg, mkdir) and run Python scripts with arguments (SKILL.md).\n
Sanitization: The instructions do not specify any sanitization or validation for timestamps, URLs, or domains extracted from the transcript before they are used in shell commands (SKILL.md).

x-post-extractor