canvas-accessibility-auditor

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data from external sources during the audit process.
  • Ingestion points: Untrusted HTML content is retrieved from Canvas pages and assignments via the get_page_content tool, and external report data is ingested via fetch_ufixit_report.
  • Boundary markers: The instructions do not define clear delimiters or "ignore embedded instructions" warnings when processing the retrieved HTML, which could allow malicious instructions embedded in a Canvas page to influence agent behavior.
  • Capability inventory: The agent has write access to course content via the edit_page_content tool and read access via get_page_content and scan_course_content_accessibility.
  • Sanitization: There is no evidence of sanitization, escaping, or validation of the external HTML content before the agent interprets it to suggest or apply accessibility fixes.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 01:12 PM