canvas-course-builder

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No direct instructions to bypass safety guidelines, reveal system prompts, or override agent behavior were found.
  • [PROMPT_INJECTION]: Evaluated the risk of indirect prompt injection as the skill processes external JSON templates and course structure data. This attack surface is mitigated by the skill's design, which requires a manual structure preview and explicit user approval before performing any write operations. 1. Ingestion points: User-provided descriptions, JSON template files, and existing course data retrieved via 'get_course_structure' (SKILL.md). 2. Boundary markers: Mandatory 'Structure Preview' step and explicit 'User Approval' step prior to execution. 3. Capability inventory: Access to Canvas creation tools including 'create_module', 'create_page', 'create_assignment', and 'create_discussion_topic'. 4. Sanitization: Relies on manual human verification of the build plan.
  • [SAFE]: No evidence of hardcoded credentials, sensitive file path access, or unauthorized network activity was identified. The skill operates through authorized MCP tool interfaces.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 01:12 PM