The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes subprocess.run in scripts/generate_schematic.py to orchestrate internal script execution for image generation tasks.
[EXTERNAL_DOWNLOADS]: The script scripts/generate_schematic_ai.py performs network communication with openrouter.ai to access generative models. This involves sending user-provided prompts and generated image data to a remote API.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its handling of untrusted clinical data. Ingestion points occur in scripts/validate_case_report.py and scripts/check_deidentification.py where medical text is processed. The skill lacks explicit boundary markers or instructional sanitization, which, when combined with its access to the Bash and Write tools, presents a low-severity security risk.

clinical-reports