Documentation

Everything you need to integrate GroundTruth into your stack.

AI Agent

CrewAI, LangChain, Strands, custom

POST /api/execute

One API call for every agent action

Policy + Verification

Fact-check claims, evaluate rules, route approvals

Tool execution

Email, Zendesk, Slack, webhooks — only if approved

Go from zero to verified AI responses in minutes.

POST /api/check — standalone fact-checking endpoint.

API key format, best practices, and error codes.

Real-time notifications when checks complete.

Copy-paste examples for Execute, Verify, and Ingest.

Usage limits and enterprise options.

Execution OS

POST /api/execute — the AI agent control plane.

Create rules to block, escalate, or allow AI actions.

Human-in-the-loop review for escalated actions.

Fact-check AI content against your knowledge base in real time.

Add the execution layer to your CrewAI agents.

Add the execution layer to your LangChain agents.

Add the execution layer to your AWS Strands agents.

Sync Help Center articles into your knowledge base.

Sync Confluence Cloud pages into your knowledge base.

Sync Notion pages into your knowledge base.