IRLFirst physical meetup — Bengaluru, Sat May 23, 4PM · RSVP on Luma
HomeToolsMCPHow It WorksStoriesPhilosophyCommunityArchitectureStar on GitHub
All Tools
G
MonitoringFreemiumOpen Source

GISKARD

Open-source evaluation and testing library for LLM agents

Apache-2.0

ABOUT

LLM agents and AI applications are vulnerable to security attacks, hallucinations, bias, and quality failures that manual testing often misses, risking data integrity and reputation. Giskard provides automated evaluation and red teaming to detect these issues during development before they reach production.

INSTALL
pip install giskard

INTEGRATION GUIDE

1. Red teaming LLM agents for prompt injection and security vulnerabilities 2. Evaluating RAG applications for hallucinations and factual accuracy 3. Detecting bias and inappropriate content in AI-generated outputs

TAGS

llm-evaluationred-teamingai-testingsecurityrag-evaluationbias-detectionmonitoringpython
Giskard — AI Tool | Agentic AI For Good