xhs-image-gen
Pass
Audited by Gen Agent Trust Hub on Mar 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Downloads the core skill logic and Python dependencies from the author's GitHub repository (github.com/PHY041/claude-agent-skills).
- [COMMAND_EXECUTION]: Runs local shell commands to manage virtual environments, install requirements via pip, and execute the generation script (generate.py).
- [PROMPT_INJECTION]: Processes user-supplied JSON content to populate image templates, creating an indirect prompt injection surface where malicious instructions could be embedded in slide content.
- Ingestion points: user-provided content_json input.
- Boundary markers: none identified in the instructions or workflow.
- Capability inventory: includes file system writes to /tmp/ and execution of subprocesses via python3.
- Sanitization: no evidence of input validation or content escaping in the documentation.
Audit Metadata