screenshot-critique
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides a mechanism for the agent to ingest and process external, untrusted visual data through a sub-agent workflow.\n
- Ingestion points: User-supplied PNG screenshots, GIF frames, and requirements documents are passed to a sub-agent in SKILL.md.\n
- Boundary markers: The skill implements security best practices by using
fork_context: falseand instructing the sub-agent to withhold project history and implementation details.\n - Capability inventory: The skill utilizes the agent's capability to spawn sub-agents (specifically of the
explorertype) to perform tasks.\n - Sanitization: The skill does not describe or implement sanitization or filtering of text that may be embedded within the images (e.g., OCR-based instruction filtering).
Audit Metadata