The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it is designed to ingest and process data from external MLB statistics providers.
Ingestion points: Data is pulled via pybaseball functions like batting_stats and statcast as described in SKILL.md and references/pull_patterns.md.
Boundary markers: The skill instructs the agent to validate schemas and coverage, but lacks explicit delimiters or specific instructions to ignore potential embedded commands in the external data results.
Capability inventory: The skill includes local script execution via scripts/smoke_load.py and file writing capabilities for Parquet artifacts.
Sanitization: No specific sanitization, filtering, or escaping of external data content is defined in the instructions.
[COMMAND_EXECUTION]: The skill provides a helper script scripts/smoke_load.py for environment verification. This script is intended to be executed to verify that the library and its dependencies are correctly installed and functional within the environment.
[REMOTE_CODE_EXECUTION]: The skill requires the installation of the pybaseball, pandas, and pyarrow packages from the official Python Package Index (PyPI). It also fetches data from well-known sports statistics providers such as Baseball Savant and FanGraphs. These operations are essential to the skill's primary functionality and utilize established public repositories and services.

pybaseball