data-labeling

Pass

Audited by Gen Agent Trust Hub on Mar 19, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from external sources like CSV files without sanitization.
  • Ingestion points: Data is loaded from files such as reviews.csv in the example scripts.
  • Boundary markers: The skill lacks explicit boundary markers or instructions to the agent to ignore any command-like text within the datasets being processed.
  • Capability inventory: The skill interacts with external APIs (Label Studio) to create projects and manage task data.
  • Sanitization: No input validation or sanitization routines are implemented for the ingested data.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 19, 2026, 08:23 AM
Security Audit — agent-trust-hub — data-labeling