viral-carousel
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (slides and captions) fetched from social media platforms, which creates a surface for indirect prompt injection attacks.
- Ingestion points: The skill ingests slide content and captions from Instagram accounts or user-provided links in the analysis phase (Instruction 4).
- Boundary markers: The instructions do not define clear delimiters or use 'ignore embedded instructions' warnings when passing external content to the model for breakdown.
- Capability inventory: The skill is capable of rendering HTML to PNG and posting content directly to Instagram using the Ootto or Composio connectors (Instruction 6).
- Sanitization: There is no evidence of filtering or sanitizing the fetched content before it is processed by the AI.
- [COMMAND_EXECUTION]: The skill uses external tools and APIs to perform its primary functions, including content retrieval and social media interaction.
- Leverages the Instagram Graph API and Composio/Ootto connectors for 'business_discovery' and posting actions.
- Instructions involve rendering slides via HTML to PNG or other design tools before posting.
Audit Metadata