remotion-component-gen
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- PROMPT_INJECTION (SAFE): No evidence of instruction override, jailbreak attempts, or system prompt extraction was found. The instructions are strictly task-oriented and instructional.\n- DATA_EXFILTRATION (SAFE): No network operations or access to sensitive local file paths (such as credentials or SSH keys) were detected. Asset references are limited to standard local project paths.\n- REMOTE_CODE_EXECUTION (SAFE): Although the skill generates TSX code, it does not download or execute untrusted scripts from external sources. Code generation follows fixed, safe templates for animation and layout.\n- DYNAMIC_EXECUTION (SAFE): The skill generates UI components based on static patterns. It does not employ unsafe deserialization, runtime compilation of untrusted code, or other dangerous dynamic loading techniques.\n- INDIRECT_PROMPT_INJECTION (SAFE): The skill ingests visual descriptions to generate code. While this represents an ingestion surface, the processing is confined to generating specific UI code within a known framework, with no evidence of malicious redirection logic.
Audit Metadata