polars

Pass

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill documents numerous methods for the agent to ingest and process data from external sources, creating a surface for potential injection attacks.
  • Ingestion points: The documentation includes functions for reading from CSV, Parquet, JSON, Excel, and SQL databases in both SKILL.md and references/io_guide.md.
  • Boundary markers: The skill does not instruct the agent on the use of delimiters or specific instructions to ignore embedded commands within the data being processed.
  • Capability inventory: The agent is empowered to read and write data to the local filesystem, network databases, and major cloud providers (AWS, Azure, and Google Cloud).
  • Sanitization: No guidance is provided for validating or sanitizing the content of the DataFrames before they are processed by the LLM or used in further system operations.
  • [EXTERNAL_DOWNLOADS]: The skill contains instructions for installing the polars package and suggests using libraries like connectorx and pyarrow. These downloads target standard public registries like PyPI, which are considered well-known and trusted sources.
  • [COMMAND_EXECUTION]: The skill uses shell commands for environment management and package installation (uv pip install), and documents patterns for executing SQL queries against external database servers.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 19, 2026, 09:10 AM
Security Audit — agent-trust-hub — polars