HomeProjectsArticlesPK BoardAbout
Explore Projects

Giskard

Active
GitHub Python Apache-2.0

Description

AI testing and vulnerability scanning platform covering LLM bias, hallucination, and security.

Key Features

  • Vulnerability scanning — Auto-detect RAG hallucination, prompt injection, toxicity, and bias
  • Multi-model — Integrations for HuggingFace, OpenAI, Anthropic, LangChain
  • Test set management — Centralized test sets and evaluations
  • CI integration — GitHub Actions / GitLab CI in one step
  • Visualization — Detailed HTML reports and team dashboard
  • Customizable — Custom tests for business needs

Use Cases

💡 Auto-test bias and hallucination before LLM rollout.
💡 Block prompt injection risks in CI.
💡 End-to-end quality regression for RAG apps.

Strengths & Limitations

✅ Strengths

  • • Actively maintained, recent updates
  • • High community interest (5.8k stars)
  • • Permissive open-source license (Apache-2.0)
  • • Established track record (4 years in production)

Tags

llm-eval vulnerability-scan quality agent-tools

Categories

📊 Observability

Quick Start

# Install
pip install giskard
# Scan an LLM
from giskard import scan
report = scan(my_llm_app)
report.to_html('report.html')
Visit GitHub

Related Projects

AxonHub

5.0k · Go
Active A

Open-source AI gateway that lets any SDK call 100+ LLMs. Built-in failover, load balancing, cost controls, and end-to-end tracing for production AI agent workloads.

agent-toolsllm-gatewayobservability +2
  • · Unified LLM gateway - call 100+ models through OpenAI and Anthropic compatible protocols
  • · Automatic failover - switch to backup providers when the primary one is down
  • · Smart load balancing - distribute requests across model instances to avoid overload
View on GitHub

Weights & Biases

11.2k · Python
Active A

Weights & Biases is an experiment tracking, visualization, and collaboration platform for ML and LLM applications, covering agent training evaluation, hyperparameter management, and model registry workflows.

observabilityexperiment-trackingmlops +2
  • · Experiment tracking — Automatically log hyperparameters, metrics, system resources, and code versions with side-by-side comparison
  • · W&B Models — Provides model artifacts registry, versioning, and promotion to production
  • · W&B Weave — LLM and agent tracing tool with prompt evaluation, conversation replay, and quality scoring
View full details → View on GitHub

Related Articles

RAGhallucination-detectionagent-evaluation

Agent Hallucination Defense: Practical Mitigation Patterns Beyond Guardrails

Why do LLM agents hallucinate? This article traces root causes and systematically reviews practical mitigation patterns: retrieval augmentation, confidence scoring, multi-agent cross-validation, forced citation backtracking, and observability with UpTrain, Giskard, RagaAI Catalyst, Comet Opik, and NVIDIA Garak.

Read more →