The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface where user-supplied inputs are incorporated into AI prompts and system commands.
Ingestion points: User-provided 'niche' and 'concept' parameters in the scripts/generator.py and scripts/generate_tiktok_viral.py files.
Boundary markers: No explicit delimiters are used to isolate user-provided data from instructional templates.
Capability inventory: The skill performs network operations via external AI APIs and executes local shell commands using FFmpeg for video processing.
Sanitization: There is no evidence of input validation or escaping for the strings processed by the pipeline.
[COMMAND_EXECUTION]: The skill constructs and executes shell commands for media manipulation.
Evidence: The skill uses FFmpeg via the ffmpeg_editor.py script for looping and compressing video content, specifically calling commands like ffmpeg -stream_loop and ffmpeg -i with dynamic parameters.
[EXTERNAL_DOWNLOADS]: The skill relies on well-known external services and their associated software packages.
Evidence: Documentation specifies dependencies on the runware and vastai-sdk Python packages, and the skill interacts with official endpoints for NVIDIA, BytePlus, Groq, and Vast.ai.

content-generator