skills/itamaker/skills/datasetlint/Gen Agent Trust Hub

datasetlint

Pass

Audited by Gen Agent Trust Hub on Oct 31, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The documentation provides instructions to install a CLI tool from the author's GitHub repository and Homebrew tap. These resources are controlled by the skill author.
  • [COMMAND_EXECUTION]: The skill executes the datasetlint binary. It includes instructions to use non-interactive subcommands to avoid execution hangs in agent environments.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external JSONL files, creating an ingestion point for untrusted data that could attempt to influence the agent via the tool's output.
  • Ingestion points: Data is read from files specified by the -train and -eval flags in the audit command.
  • Boundary markers: No delimiters or instructions are used to separate data from instructions within the JSONL records.
  • Capability inventory: The skill facilitates file reading for auditing purposes and shell command execution.
  • Sanitization: The tool validates the JSON schema and data types but does not implement sanitization for the textual content of the data fields.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 31, 2026, 08:38 AM