canva
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell scripts (scripts/canva.sh and scripts/canva-auth.sh) to perform API operations using system tools such as curl, jq, and openssl.
- [EXTERNAL_DOWNLOADS]: The skill's documentation includes instructions for downloading and installing the skill from GitHub or using npx, which involves fetching external code.
- [CREDENTIALS_UNSAFE]: The skill manages sensitive Canva API credentials (Client ID and Secret) through environment variables and persists OAuth tokens in a local configuration file (~/.canva/tokens.json).
- [PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface by interpolating user-provided text directly into Canva template autofill payloads. 1. Ingestion points: scripts/canva.sh (autofill command). 2. Boundary markers: Absent. 3. Capability inventory: Network operations via curl to api.canva.com. 4. Sanitization: Absent.
Audit Metadata