The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied research data to generate visualizations, creating a potential surface for indirect injection.
Ingestion points: User-provided research topics, element descriptions, and relationship definitions defined in Step 1 of SKILL.md.
Boundary markers: The instructions lack specific delimiters or instructions to the agent to treat interpolated user data as non-executable text.
Capability inventory: The skill has access to Bash for running Python scripts, Write for saving output files, and the Excalidraw MCP for rendering diagrams.
Sanitization: There are no explicit validation or sanitization steps identified in the instructions or scripts for the user-provided labels and descriptions.
[DYNAMIC_EXECUTION]: The skill dynamically generates and executes code to produce visualizations.
Script generation: Step 8 of SKILL.md provides a Python template for generating matplotlib charts based on research data, and the skill invokes scripts/generate_prisma.py to render PRISMA diagrams.
Interactive output: The skill constructs JSON structures at runtime to invoke the Excalidraw:create_view tool for interactive diagram editing.

tw-research-viz