The Agent Skills Directory

[SAFE]: All components of the skill, including the instructions and the provided Python script, are aligned with the stated purpose of data visualization and do not perform any unauthorized or suspicious actions.
[EXTERNAL_DOWNLOADS]: The skill's static rendering mode utilizes the AntV library from unpkg.com, which is a recognized and trusted content delivery network for web dependencies. This is a standard method for including open-source libraries in web-based visualizations.
[COMMAND_EXECUTION]: The agent is instructed to run a local script, gif_chart.py, to generate animations. The script uses well-established libraries (matplotlib, numpy, Pillow) and implements path validation for outputs, ensuring that file operations are restricted to the intended project directories.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user input to generate infographic content.
Ingestion points: User-provided text, CSV, and JSON data are used to populate chart fields in SKILL.md and references/antv-templates.md.
Boundary markers: There are no specific delimiters to isolate user-supplied strings from the rendering logic.
Capability inventory: The skill can write files to the outputs/ directory and execute a local Python script.
Sanitization: No explicit sanitization or filtering of user data was identified, but the execution environment for visualizations is localized and presents minimal risk to the agent's core logic.

infographic