Skills
Navigate
skills/ml4t/skills/ml4t-point-in-time/Gen Agent Trust Hub

ml4t-point-in-time

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill demonstrates ingesting data from external Parquet files, which is a potential surface for indirect prompt injection if the data source is untrusted.
  • Ingestion points: pl.read_parquet("prices.parquet") and pl.read_parquet("fundamentals.parquet") in SKILL.md.
  • Boundary markers: No delimiters or ignore instructions are present in the code examples.
  • Capability inventory: The code demonstrates local file read operations.
  • Sanitization: No sanitization or validation of the input data is described.
  • [COMMAND_EXECUTION]: The provided code snippets execute file system operations to load datasets for processing.
  • Evidence: prices = pl.read_parquet("prices.parquet") in SKILL.md.
  • Context: These operations are standard for the skill's primary purpose of financial data processing.
  • [EXTERNAL_DOWNLOADS]: The skill references the polars library and the vendor's own ml4t-data package.
  • These are documented as standard dependencies for the provided implementation examples and follow expected vendor resource patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 01:50 PM