The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external configuration files (config.json and config.example.json) to define image prompts and audio timing. This data is interpolated into media processing workflows without sanitization or boundary markers, creating a surface where malicious instructions in the input files could potentially influence agent behavior.\n
Ingestion points: scripts/config.example.json and the user-provided config.json containing keyframe_prompt and audio metadata.\n
Boundary markers: Absent; the skill does not define delimiters to isolate untrusted configuration data from system instructions.\n
Capability inventory: The skill utilizes ffmpeg for video concatenation and ken-burns effects, and PIL (Pillow) for image compositing, involving multiple subprocess calls across the pipeline.\n
Sanitization: Absent; the skill does not describe any validation or escaping of input values before they are used in media processing commands.\n- [COMMAND_EXECUTION]: The skill provides and instructs the agent to execute multiple shell commands for media processing and tool installation.\n
Evidence: PIPELINE.md and README.md contain bash code blocks for ffmpeg operations such as clipping, muxing, and applying filters.\n
Evidence: The skill references the execution of a Python script (caption.py) from a related capability and utilizes npx for the installation and execution of vendor-provided tools.

render-editorial-motion-podcast