Examples
Full, runnable examples for common VeriProof integration patterns. Each example is self-contained and can be copied directly into your project.
All examples require a VeriProof API key. Set VERIPROOF_API_KEY and VERIPROOF_APPLICATION_ID as environment variables, or pass them explicitly via the options object.
TypeScript / Node.js
Trace generateText and streamText calls with automatic span capture using @veriproof/sdk-instrumentation-ai-sdk.
Record a full decision session with risk levels, guardrail results, and EU AI Act metadata using the fluent VeriproofSession API.
Auto-instrument LangChain.js chains and agents with @veriproof/sdk-instrumentation-langchainjs.
Python
Wrap a Pydantic AI agent in a VeriProof governance session with intent, risk level, and outcome recording.
Pydantic AIInstrument a LangGraph state-machine chatbot with per-node span capture and session-level governance metadata.
LangGraph ChatbotOrchestrate three sequential agents (classify → assess → decide) under a single VeriProof session.
Multi-Turn Workflow.NET
Auto-instrument Microsoft Semantic Kernel function invocations and model calls using VeriproofKernelFilter.
Manual Activity tracing with full governance annotations for a multi-step loan decisioning workflow.
Emit agent_run and chat_completion spans across multiple cooperating agents using AutoGenOTelAdapter.
What’s in each example
Every example follows the same structure:
- Prerequisites — what to install
- Environment setup — the minimum
.envneeded - Complete runnable code — copy-paste ready
- What you’ll see in VeriProof — the spans and governance metadata that appear in the dashboard
- Next steps — related SDK reference and guides