dreamina-canvas-cli-text2voice
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill takes user-provided narration text and interpolates it into shell commands for the
dreamina-canvasCLI. If the input contains shell metacharacters, it could lead to unintended command execution. - Ingestion points: Narrated text is captured in parameters like
<narration>and<text>inSKILL.mdandreferences/parameter-reference.md. - Boundary markers: The prompt text is enclosed in double quotes within the shell command template (e.g.,
--prompt "<narration>"). There are no specific instructions for the agent to sanitize the input against shell injection. - Capability inventory: The skill executes shell commands via the
dreamina-canvasCLI to create audio nodes. - Sanitization: The instructions do not specify any validation or sanitization steps for the narration text before it is passed to the shell.
- [COMMAND_EXECUTION]: The primary function of the skill involves generating and executing shell commands using the
dreamina-canvastool to manage audio assets and drafts. - [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install additional skill components from the vendor's ecosystem using
npx. - Evidence:
SKILL.mdcontains the rule:npx skills add full-aigc-skills/dreamina-skills --skill <skill-name>.
Audit Metadata