space-ai-article-illustration
Pass
Audited by Gen Agent Trust Hub on Jul 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script (
scripts/generate_image.py) to perform the image generation and file management tasks. This execution is the intended primary function of the skill. - [EXTERNAL_DOWNLOADS]: The skill interacts with external APIs at
api.labnana.comandgenerativelanguage.googleapis.com(Google Gemini) to fetch generated image data. These are legitimate endpoints required for the skill's operation. - [PROMPT_INJECTION]: The skill analyzes user-supplied Markdown content to generate image prompts, which introduces a surface for indirect prompt injection.
- Ingestion points: Untrusted article content is read from the user's files (SKILL.md, Phase 1).
- Boundary markers: The skill includes a clear safety boundary by requiring the agent to display all generated prompts and wait for explicit user confirmation before proceeding to the generation phase (SKILL.md, Phase 1, Step 5).
- Capability inventory: The agent uses a subprocess call to run a local script with network access and file-write capabilities (SKILL.md, Phase 2).
- Sanitization: The workflow relies on the human-in-the-loop confirmation step to validate the generated prompts and commands.
Audit Metadata