verify
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process external data from running applications and browser pages to verify behavioral changes.
- Ingestion points:
SKILL.mdinstructions (Sections 1 and 2) explicitly require the agent to open and observe application output, pages, and browser windows. - Boundary markers: Absent; there are no instructions to treat content found in the running application as untrusted or to ignore embedded instructions.
- Capability inventory: Following these instructions requires the agent to have UI interaction capabilities such as
screenshot,browser, orcomputer_use. - Sanitization: Absent; the skill does not prescribe any filtering, validation, or escaping of the content ingested from the external application.
Audit Metadata