design-lint
Pass
Audited by Gen Agent Trust Hub on Jul 25, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill includes a surface for indirect prompt injection due to its core function of processing untrusted content from local files and external web pages.
- Ingestion points: The instructions in SKILL.md (Phase 1 and Phase 2) direct the agent to analyze local file paths and visit external URLs supplied by the user.
- Boundary markers: There are no explicit delimiters or instructions provided to ensure the agent ignores potential malicious instructions embedded within the CSS comments or HTML content of the targets being audited.
- Capability inventory: The skill utilizes file read operations (grep), network access to open URLs, and browser capabilities for screenshotting.
- Sanitization: No sanitization or validation of the external content is performed before the agent processes and judges the data.
Audit Metadata