25 docs tagged with "Reference"

Configure the OpenBox Claude Code integration: hook commands, environment variables, observe vs. enforce modes, and per-developer privacy controls.

All configuration options for the OpenBox Deep Agents SDK: environment variables, middleware parameters, and tool type mapping.

Configure the OpenBox LangGraph SDK: environment variables, handler parameters, governance behavior, and instrumentation options.

Configure the OpenBox plugin: Set API keys, define policies, control runtime behavior - no infrastructure changes needed.

Deep Agents demo architecture: middleware hooks lifecycle, governance event flow, subagent dispatch, and OpenTelemetry instrumentation.

See governance in action: Explore demo architecture with Temporal workflows, custom policies, and runtime enforcement patterns.

Handle OpenBox governance errors in DeepAgents: exception types, try/except patterns, and debugging guidance.

Handle OpenBox governance errors in LangChain agents: exception types, try/except patterns, and debugging guidance.

Handle OpenBox LangGraph SDK errors: governance blocks, guardrail violations, HITL rejections, and network failures.

Handle OpenBox startup failures, runtime verdicts, approval states, and guardrail validation errors in Mastra.

Handle OpenBox governance errors in n8n: node error types, Continue On Fail behavior, and debugging guidance.

Handle OpenBox errors gracefully: Manage policy violations, trust failures, and network issues with retry logic and fallback patterns.

Understand how DeepAgents invocations, model calls, tools, subagents, and telemetry are represented in OpenBox.

Understand how LangChain agent runs, model calls, tools, and telemetry are represented in OpenBox.

Understand how LangGraph graph runs, tools, LLM calls, signals, and telemetry are represented in OpenBox.

Understand how Mastra workflows, tools, agents, and telemetry are represented in OpenBox.

Understand how n8n agent runs, model calls, tools, and telemetry are represented in OpenBox.

OpenBox terminology explained: Definitions for trust scores, attestation, policies, governance decisions, and AI agent concepts.

API reference for the OpenBox OpenRouter SDK: govern callModel and tools, read routing summaries, and locate run receipts.

Developer-facing plumbing for OpenBox itself: subscribe to governance events via webhooks, manage platform API keys, and roll out features with flags.

Every attribute the OpenRouter SDK seals onto a governed model call: provider, region, cost, failover trail, and the requested-versus-actual comparisons.

The vocabulary of gateway routing: allowlists, routing decisions, routing records, and the evidence they produce.

OpenBox plugin reference for Temporal Python: governance, policy enforcement, trust scoring, and monitoring via a single plugin.

Diagnose missing routing provenance, an empty Provenance tab, calls that fail closed, and receipts that will not verify.

Help AI assistants understand your setup: Enable LLMs to read your governance policies, agent config, and documentation automatically.