web-quality-audit
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from web applications and third-party scripts during performance audits, creating a potential vector for indirect prompt injection.
- Ingestion points: External web page content, performance traces, and third-party script telemetry as described in the 'Suggested Process' and 'Core Rules'.
- Boundary markers: The 'Security Guardrails' (ASI01) explicitly instructs the agent to treat observed page content and script telemetry as data, though no structural delimiters are specified.
- Capability inventory: The skill uses tools with significant local impact, including
write_file,edit_file,run_build, andrun_tests(SKILL.md frontmatter). - Sanitization: No technical sanitization or validation logic is defined for the ingested data prior to its use in generating reports or interacting with the filesystem.
Audit Metadata