dreamina-canvas-cli-text2image
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that ingests user-supplied prompts for interpolation into shell commands, creating a potential surface for indirect prompt injection attacks. \n
- Ingestion points: User input is used for the
<prompt>placeholder inSKILL.md(Step 3) when creating image nodes. \n - Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the prompt input. \n
- Capability inventory: The skill leverages shell command execution through the
dreamina-canvasCLI as documented inSKILL.mdandreferences/workflow-patterns.md. \n - Sanitization: No sanitization or escaping of the prompt data is mentioned in the instructions. \n- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing additional components from the vendor's repository. \n
- Evidence:
SKILL.mdsuggests usingnpx skills add full-aigc-skills/dreamina-skillsto acquire missing functionalities.
Audit Metadata