canva-core-workflow-b
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external content from the Canva API, which presents a surface for indirect prompt injection. 1) Ingestion points: Step 4 in SKILL.md specifies reading template and design datasets from the Canva API. 2) Boundary markers: The instructions do not provide explicit delimiters or prompts for the agent to ignore instructions embedded in the external data. 3) Capability inventory: The skill utilizes Write and Edit tools as part of its operations. 4) Sanitization: The instructions include validation steps for resource ownership and malware policies (Step 2), and mapping validation (Step 4), but do not explicitly detail sanitization for prompt injection.
- [EXTERNAL_DOWNLOADS]: The skill references API documentation and resource guides from Canva's official developer domain (canva.dev).
Audit Metadata