yt-ideation
Pass
Audited by Gen Agent Trust Hub on Apr 4, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill contains deceptive or inconsistent metadata. While the primary sections describe a YouTube ideation process, the 'Instructions', 'Prerequisites', and 'Error Handling' sections reference 'ORM configuration', 'ORM environment', and 're-authenticating with ORM'. This discrepancy could cause an agent to perform unintended actions on database configurations while the user expects content brainstorming.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its reliance on processing untrusted external data without proper boundaries.
- Ingestion points: The skill reads 'niche-analysis.json' and 'niche-report.md' (Step 2) and retrieves content from external sources like Reddit and YouTube comments (Method 4).
- Boundary markers: There are no instructions or delimiters provided to ensure the agent ignores instructions embedded within the processed research data or audience comments.
- Capability inventory: The skill is authorized to use 'WebSearch', 'Read', 'Write', and 'Task' tools, which provides a significant capability surface if malicious instructions are ingested.
- Sanitization: No validation, escaping, or filtering of the external input files or web content is implemented.
Audit Metadata