The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion and processing of DataFrames, creating a surface for potential indirect prompt injection if the processed data contains malicious instructions.
Ingestion points: Data enters the agent's context through schema.validate(df) calls and the @pa.check_types decorator used in pipeline steps.
Boundary markers: No explicit instruction delimiters or warnings to ignore embedded instructions are present in the provided examples.
Capability inventory: The skill includes capabilities to write to the local file system via pandera.io.to_yaml and generate Python scripts with pandera.io.to_script.
Sanitization: No specific sanitization or escaping of DataFrame content is mentioned prior to validation.
[DYNAMIC_EXECUTION]: The skill documents the ability to generate executable Python code from data schemas.
Evidence: The pandera.io.to_script(schema, "./schema_definition.py") function is used to create a Python script representing the inferred or defined schema.
[EXTERNAL_DOWNLOADS]: The skill references the installation of the Pandera library from a public registry.
Evidence: Documentation in references/checks-and-validation.md suggests using pip install 'pandera[hypotheses]' to access full functionality.

pandera