Solutions

Two ways to test an LLM agent with the open-source library, depending on whether you want the test cases generated or written.

Scan Vulnerabilities Describe your agent and let the scan red team it. Install, wrap, run, read the report.

Check Agentic Systems Write the assertions yourself: state a rule in plain English, collect scenarios into a suite, run it in CI.

Each page is a quickstart. The full documentation lives in Giskard Scan and Giskard Checks. Most teams end up running both: the scan to discover vulnerabilities they had not thought of, checks to lock in the behavior they already care about.