dreamina-canvas-cli-text2audio
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill operates by executing the
dreamina-canvasCLI tool to manage project nodes and music generation tasks. Subcommands includenode create,model list, andoperation wait. - [INDIRECT_PROMPT_INJECTION]: The skill accepts music descriptions from the user and interpolates them directly into the
--promptargument of shell commands. This creates a surface for potential indirect prompt injection attacks. - Ingestion points: Music prompts provided by users in
SKILL.mdandexamples/happy-path.md. - Boundary markers: The instructions do not define specific delimiters or instructions to the model to ignore embedded commands within the user prompt.
- Capability inventory: The skill has the ability to trigger music generation and interact with remote services via the
dreamina-canvasCLI. - Sanitization: There is no explicit mention of shell character escaping or input validation for the user-supplied prompt strings.
- [EXTERNAL_DOWNLOADS]: The skill instructions provide a command to install additional skills from the same vendor (
npx skills add full-aigc-skills/dreamina-skills). This is a standard dependency within the vendor's ecosystem and is documented as part of the setup process.
Audit Metadata