autodiscovery

Pass

Audited by Gen Agent Trust Hub on May 18, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill ensures the environment is correctly configured by installing the asta-plugins package from the official allenai GitHub repository. This installation is performed using the uv tool and targets a specific version (v0.17.0) from the vendor's repository.
  • [COMMAND_EXECUTION]: Utilizes standard shell commands and the asta CLI to perform administrative tasks, such as listing runs, checking credit balances, and submitting experiments. It also uses benign data inspection commands like head and a Python JSON-parsing one-liner to help the user understand their datasets.
  • [SAFE]: No instances of prompt injection, obfuscation, or unauthorized data exfiltration were found. The skill manages authentication through the official asta auth login process and interacts only with user-specified data files and the vendor's verified platform.
Audit Metadata
Risk Level
SAFE
Analyzed
May 18, 2026, 09:12 AM
Security Audit — agent-trust-hub — autodiscovery