manim-composer
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from user-provided concepts and external information gathered during its research phase via web search. This establishes an ingestion surface where malicious instructions could potentially be embedded in external content. However, the risk is mitigated as the skill's primary output is a descriptive markdown document for planning purposes, rather than executable code.
- Ingestion points: User-provided animation concepts (SKILL.md) and external information retrieved during the web search research phase (SKILL.md Phase 1).
- Boundary markers: No explicit delimiters or instructions to ignore instructions embedded in search results are provided.
- Capability inventory: The skill utilizes web search but does not include scripts for file system modification, command execution, or network exfiltration beyond standard search tasks.
- Sanitization: No specific sanitization or validation logic is defined for the content retrieved from external sources.
Audit Metadata