The Agent Skills Directory

[COMMAND_EXECUTION]: The wrapper script generate_infographic.py uses subprocess.run() to execute the internal logic script generate_infographic_ai.py. The command is constructed using a list of arguments, which is a secure practice that prevents shell injection attacks.
[DATA_EXFILTRATION]: The skill communicates with the OpenRouter API (openrouter.ai) to generate infographic images and perform quality reviews. This network activity is the primary function of the skill and is directed at a well-known service. The script securely manages API keys by retrieving them from environment variables or a local .env file.
[PROMPT_INJECTION]: The skill ingests user-provided text prompts and interpolates them into instructions for external AI models (Perplexity Sonar and Gemini). While it lacks strong boundary markers (e.g., XML tags or unique delimiters) around the user input, the logic is consistent with standard AI tool implementation and contains no malicious instructions or bypass attempts.

infographics