Product / UX / Safety & Guardrails

Giskard

An open-source Python library for testing and red-teaming LLM and agentic systems. Giskard Scan automatically probes applications for vulnerabilities (prompt injection, harmful output, hallucination), and Giskard Checks runs scenario-based evaluations, including multi-turn agent validation.

Giskard, Apache-2.0, Python; ~5.5K GitHub stars (June 2026). Scan (automated red-teaming) + Checks (scenario evals). Verified live June 2026.

Openness

5 high confidence

5.0

license: Apache-2.0(OSI)
source: public
self-host: yes
service: none(per the repo)
core-gated: ungated

Fully open source: Apache-2.0 Python library, self-hostable; no commercial tier mentioned on the repo.

https://github.com/Giskard-AI/giskard recorded 2026-06-29
Apache-2.0; ~5.5k stars; automated scan + scenario checks; adversarial/injection testing

Adoption

3 low confidence

3.0

~5.5K GitHub stars (June 2026); stars-only proxy, capped per methodology.

https://github.com/Giskard-AI/giskard recorded 2026-06-29
~5.5k stars (June 2026)

Capability

4 medium confidence

4.0

Combines automated red-team scanning with structured evaluation; broad coverage for both safety and reliability.

https://github.com/Giskard-AI/giskard recorded 2026-06-29
Scan + Checks; adversarial, injection, multi-turn agent validation

Unchanged since 2026-08-01 (last edited, not re-checked)