The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external research via Perplexity Sonar and interpolates it into AI prompts for infographic generation without explicit boundary markers or sanitization.
Ingestion points: In scripts/generate_infographic_ai.py, the research_topic function retrieves content from an external API, which is then merged into the prompt in _enhance_prompt_with_research.
Boundary markers: Absent; external content is appended to the prompt using a plain text header.
Capability inventory: The skill has the capability to write files (PNG images and JSON logs) to the local file system and perform outbound network requests via the requests library.
Sanitization: No input validation or escaping is performed on the retrieved research content before it is processed by the image generation model.
[COMMAND_EXECUTION]: The entry point scripts/generate_infographic.py uses the subprocess module to execute the secondary generation script.
Evidence: The script builds a command list and executes scripts/generate_infographic_ai.py using subprocess.run. The command is constructed as a list without a shell, which is a standard execution pattern but involves dynamic local script invocation.
[EXTERNAL_DOWNLOADS]: The skill communicates with the OpenRouter API to perform topic research, generate images, and conduct quality reviews.
Evidence: The script scripts/generate_infographic_ai.py makes network requests to openrouter.ai endpoints to access models like google/gemini-3.1-pro-image-preview and perplexity/sonar-pro. This is a documented feature and utilizes well-known AI service providers.

alterlab-infographics