critique-accessibility
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted HTML and Markdown data provided by the user for accessibility auditing. This constitutes an indirect prompt injection attack surface.
- Ingestion Points: Artifact content or file paths are passed to
scripts/checks.pyand thecritique-criticsubagent. - Boundary Markers: The skill does not explicitly define boundary markers for the artifact content in its instructions.
- Capability Inventory: The skill performs structural analysis and CSS resolution using Python's
HTMLParser. It generates structured JSON findings but does not execute the artifact content. - Sanitization: Artifact content is parsed as data by
scripts/checks.py. The skill relies on standard library parsing and a shared runner for processing.
Audit Metadata