screenshot-analyzer
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process user-provided screenshots, which serves as an entry point for untrusted data. Malicious text embedded within an image could attempt to override the instructions of the synthesis or reviewer agents.
- Ingestion points: User-supplied screenshot files are read and processed in Phase 1 and Phase 2 (SKILL.md).
- Boundary markers: The skill does not implement explicit delimiters or "ignore embedded instructions" warnings when passing data between the UI, Interaction, and Business analyzers.
- Capability inventory: The skill utilizes the Task tool for coordination and performs local filesystem writes to the
docs/plans/directory. - Sanitization: The instructions do not define any sanitization or validation steps for content extracted from images before it is interpolated into the synthesizer and reviewer prompts.
Audit Metadata