logo-design-guide
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
infshCLI tool via the Bash(infsh *) tool to run remote AI models. These commands are used for generating images, upscaling, and managing user sessions on the inference.sh platform. Evidence found in the workflow examples in SKILL.md. - [EXTERNAL_DOWNLOADS]: The documentation references installing additional functionalities using
npxfor vendor-specific skills (e.g.,@inference-sh/skills). These resources belong to the same vendor context as the skill author and are used to extend agent capabilities. - [PROMPT_INJECTION]: The skill template interpolates user-controlled data into prompts for the image generation models. While this presents an indirect prompt injection surface, it is a necessary part of the logo design functionality.
- Ingestion points: User-provided logo descriptions in the
promptfield of theinfsh app runcommands in SKILL.md. - Boundary markers: Absent. No explicit delimiters or instructions to ignore embedded commands are present in the provided templates.
- Capability inventory: The
Bash(infsh *)tool provides the ability to execute remote applications and perform network operations as shown in SKILL.md. - Sanitization: Absent. User input is interpolated directly into the tool's JSON input payload.
Audit Metadata