brand-onboarding-codex
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from external websites and social profiles to build a brand profile. Malicious instructions embedded in these external sources could potentially influence the agent's behavior during the analysis phase.
- Ingestion points: External website copy, service/product pages, blogs, and competitor pages gathered via research tools like Firecrawl or Playwright MCP as described in SKILL.md.
- Boundary markers: The instructions do not define specific delimiters or "ignore instructions" markers to isolate the untrusted external data from the agent's core logic.
- Capability inventory: The skill's primary capability is writing structured documentation to the local filesystem (e.g.,
context/brand-style.md). - Sanitization: No specific sanitization, validation, or filtering of the gathered external content is performed before the agent processes it.
Audit Metadata