The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the installation of content from an external community GitHub repository (github.com/oil-oil/oil-motion) using the npx skills add command.
[REMOTE_CODE_EXECUTION]: The skill instructions specify that the agent will run pip install -r scripts/requirements.txt from the downloaded repository at runtime. This pattern involves executing dependency manifests from an untrusted community source.
[COMMAND_EXECUTION]: The installation process and the detection logic in the 'Decision tree' involve executing shell scripts to check for the existence of files and install software via npx and pip.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of AI-generated or existing video and image sequences.
Ingestion points: User-provided image and video sequences processed for animation (SKILL.md).
Boundary markers: None specified for delimiting instructions within media metadata or associated prompts.
Capability inventory: The skill executes shell commands (pip, ffmpeg), performs network operations (ZenMux API), and writes animation assets to the local file system.
Sanitization: No evidence of sanitization or validation of the input media content or metadata before processing.

oil-motion