article-image

Pass

Audited by Gen Agent Trust Hub on Jul 25, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONCREDENTIALS_UNSAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes several Bash and Python scripts (gen-image.sh, gpt-image-2-gen.sh, extract_image.py) to interface with the Codex CLI and PicGo tool for image generation and extraction.
  • [EXTERNAL_DOWNLOADS]: The gen-image.sh script includes logic to automatically install the picgo package from official Node.js registries (via npm, pnpm, yarn, or bun) if it is missing from the environment.
  • [CREDENTIALS_UNSAFE]: To facilitate image uploads, the skill synchronizes configuration from the PicGo GUI (~/Library/Application Support/picgo/data.json) to the CLI configuration (~/.picgo/config.json). While this is necessary for the tool's function, it involves reading and writing configuration files that typically contain sensitive API tokens for image hosting services.
  • [DATA_EXFILTRATION]: The skill is designed to upload generated image files to external image hosting providers as configured by the user via the picgo tool.
  • [SAFE_PRACTICE]: The extract_image.py script includes a strict validation function for output paths, preventing the skill from writing files to sensitive system directories (e.g., /etc, /bin, /usr) and restricting file extensions to common image formats.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 25, 2026, 07:58 AM
Security Audit — agent-trust-hub — article-image