ais-bench

Pass

Audited by Gen Agent Trust Hub on May 19, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches its core benchmark suite from the official AISBench project repository on GitHub. These operations are essential for the tool's primary functionality and utilize expected infrastructure for an open-source research tool.\n- [COMMAND_EXECUTION]: Utility scripts in the scripts/ directory automate environment verification and the execution of benchmarking tasks. These scripts use shell commands to manage configurations and launch performance tests, which is standard behavior for this type of evaluation suite.
Audit Metadata
Risk Level
SAFE
Analyzed
May 19, 2026, 06:27 AM
Security Audit — agent-trust-hub — ais-bench