yc-pitch-deck
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEMETADATA_POISONINGINDIRECT_PROMPT_INJECTION
Full Analysis
- [METADATA_POISONING]: The skill documentation includes a 'Skill Boundaries' section that incorrectly describes video production capabilities (videography, video editing), suggesting a copy-paste error from an unrelated skill. Additionally, there are conflicting author names ('ClawFu' and 'MKTG Skills') within the file, both of which differ from the declared skill author ('roohe').
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data without sufficient guardrails.
- Ingestion points: User-supplied startup descriptions and metrics in the 'How to Use' section.
- Boundary markers: No delimiters or specific instructions for the agent to ignore embedded commands in user input are provided.
- Capability inventory: The skill does not have access to risky tools like network operations or shell execution.
- Sanitization: No input validation or filtering mechanisms are defined.
Audit Metadata