The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/generate_schematic.py invokes a secondary script scripts/generate_schematic_ai.py using subprocess.run(). This call uses a list-based argument structure which is resistant to shell injection and does not invoke a shell environment.
[EXTERNAL_DOWNLOADS]: The scripts scripts/generate_image.py and scripts/generate_schematic_ai.py interact with the openrouter.ai API to perform image generation. This is an established service and the usage is consistent with the skill's primary research and writing purpose.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied prompts for figure generation and interpolates them into instructions for an AI model. 1. Ingestion points: args.prompt in scripts/generate_image.py and scripts/generate_schematic_ai.py. 2. Boundary markers: Absent. 3. Capability inventory: Local script execution via subprocess and network access to an AI provider. 4. Sanitization: Absent. The security risk is assessed as low as the interpolated data is used exclusively for generating visual figures.

scientific-writing