manim-animator
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill generates Python code using the Manim animation library and executes it locally to render video files. This is the primary function of the skill.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided explanations to generate executable code, creating a surface where malicious instructions could potentially be embedded in the input to influence the agent's code generation.
- Ingestion points: User-provided conceptual explanations and narration text (SKILL.md).
- Boundary markers: None specified to separate user input from system instructions.
- Capability inventory: Python script execution, environment inspection (Python, Manim, FFmpeg, LaTeX), and external API interaction via vendor-owned sandbase_* tools (SKILL.md, references/sandbase-api-map.md).
- Sanitization: No explicit instructions for sanitizing user-provided content before interpolation into code.
- [COMMAND_EXECUTION]: The skill performs environment checks (Python, FFmpeg, LaTeX) and executes rendering commands to produce video output, including capturing exit statuses and logs.
Audit Metadata