remotion-expert

Pass

Audited by Gen Agent Trust Hub on Oct 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill documents patterns for AI-driven video modification where natural language instructions are mapped to video properties, creating a surface for adversarial content in processed data to influence agent behavior.
  • Ingestion points: Instructions are passed to aiUpdateHandler in SKILL.md to modify currentProps.
  • Boundary markers: None identified in the provided code snippets or instructions.
  • Capability inventory: The skill uses fetch in SKILL.md and references/compositions-props.md to retrieve external data based on properties.
  • Sanitization: No explicit sanitization or validation of the input instructions is demonstrated beyond the conceptual handler.
  • [COMMAND_EXECUTION]: The Quick Start section in SKILL.md provides shell commands for project scaffolding using the bun package manager.
  • [EXTERNAL_DOWNLOADS]: The skill includes code examples that fetch data or assets from external, non-whitelisted placeholder domains such as api.v2.com and example.com in SKILL.md, references/compositions-props.md, and references/media-assets.md.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 17, 2026, 08:22 AM