Approvals and Guardrails
How OpenBox verdicts, approvals, and guardrails are enforced at runtime in LangGraph.
LangGraph state machine integration
View all tagsHow OpenBox verdicts, approvals, and guardrails are enforced at runtime in LangGraph.
Configure the OpenBox LangGraph SDK: environment variables, handler parameters, governance behavior, and instrumentation options.
Handle OpenBox LangGraph SDK errors: governance blocks, guardrail violations, HITL rejections, and network failures.
Understand how LangGraph graph runs, tools, LLM calls, signals, and telemetry are represented in OpenBox.
Add OpenBox governance to LangGraph agents: wrap your compiled graph for real-time policy enforcement, guardrails, and human-in-the-loop approvals.
Govern a LangGraph application with OpenBox: install the SDK, wrap your compiled graph, configure identity, and verify live runs.
The LangGraph concepts that matter when adding OpenBox governance to a graph-based agent.
Developer reference for governing LangGraph agents with OpenBox: graph handler, policy enforcement, guardrails, HITL, and tool classification.
See what operational telemetry the OpenBox LangGraph SDK captures and how it appears in OpenBox.
Common LangGraph SDK integration issues and how to diagnose them quickly.
Add OpenBox governance to an existing LangGraph application without changing the graph topology.