ideate-notes
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from multiple external sources, creating a significant attack surface for indirect prompt injection.
- Ingestion points: As described in SKILL.md (Steps 1, 2, and 3), the agent fetches transcripts and comments from YouTube, content from Substack archive pages, and results from general web searches for trends.
- Boundary markers: The instructions do not specify any delimiters or safety prompts to ensure the agent ignores potential instructions embedded within the ingested external content.
- Capability inventory: The skill has the capability to write the results to the local filesystem (./substack/notes/ideas.md and ./substack/notes/processed-log.md) and utilizes YouTube MCP tools for channel interaction.
- Sanitization: There is no evidence of a sanitization or filtering process for the content of YouTube comments or search results before they are passed to the ideation foundation skill.
Audit Metadata