The Agent Skills Directory

[SAFE]: The skill provides standard documentation and code examples for using the well-known Matplotlib library. The instructions are consistent with legitimate visualization workflows.
[COMMAND_EXECUTION]: The skill mentions executing helper scripts (scripts/plot_template.py and scripts/style_configurator.py) via the Python interpreter. These scripts are described as internal components of the skill intended for user reference.
[EXTERNAL_DOWNLOADS]: The skill includes links to official Matplotlib documentation, galleries, and tutorials on matplotlib.org and GitHub. These are well-known technology domains and are considered safe sources.
[INDIRECT_PROMPT_INJECTION]: The skill demonstrates functions that process data and strings for plot content, such as titles and labels. This constitutes an ingestion surface where untrusted data could be used to influence output if processed by an agent.
Ingestion points: Data dictionaries and title parameters used in plotting code blocks within SKILL.md.
Boundary markers: None present; titles and labels are interpolated directly into plotting functions.
Capability inventory: The skill involves writing files to the local system (plt.savefig) and executing local scripts.
Sanitization: No input validation or sanitization for plot text (titles, labels) is mentioned in the instructions.

matplotlib