Documentation

Documentation for setup and production operations.

Use this hub to move from initial integration to enforced execution control, with identity governance, evidence exports, and operational runbooks.

Start here

Recommended adoption path for new teams:

1. Define authority and policy

Set execution scopes, approvals, and role mappings. Start with Policy Model and Role Mapping.

2. Integrate runtime

Prefer the unified adapter contract (HTTP/Webhook) for new runtimes. OpenClaw and Temporal remain compatibility paths. For the production Top 20 set, use Adapter Directory.

3. Run observe-only validation

Simulate decisions, review DENY reasons, and tune policy before blocking execution.

4. Enforce and operate

Enable ALLOW/DENY gates, evidence exports, and ongoing ops checks.

Setup Any Agent Runtime

Use this path for any agent framework or automation system.

1. Connect tenant + auth

Set API base, tenant ID, and token in Console setup.

2. Implement runtime adapter

Use HTTP/Webhook adapter: register runtime, call intent before side effects, send heartbeat.

3. Pick your framework mapping

Use targeted guides for LangGraph, Semantic Kernel, Bedrock, Vertex ADK, Automation, or Worker Queue. Keep OpenClaw and Temporal compatibility paths if already deployed.

4. Validate fail-closed behavior

Use fail-closed requirements and API contract checks before rollout.

5. Turn on operations + evidence

Run deployment checks, monitor via integration runtimes page, and enable evidence exports.

Core concepts

How to use

Runbook-style guides for the most common governance operations.

High-value operations

Documentation library

Getting started

Connect a runtime, run observe-only, then enforce.

Console usage

Connect your tenant, review evidence, approve changes, export packs.

Policy model

Scopes, approvals, budgets, and lifecycle.

Evidence model

Decision records, audit IDs, and lineage.

Identity governance

SSO, SCIM, session policy, and change queue.

SSO setup

Provider registration, metadata validation, and session issuance.

SCIM sync

User/group provisioning and drift reconciliation.

Google Workspace setup

Configure Google OIDC, map roles, and sync users/groups into Gvner.

Agent governance

Inventory, baselines, anomalies, enforcement, and bundles.

Operational integrity

Health alerts, integrity checks, SLA reporting, incidents.

Change approval workflow

Request, diff, approve, apply, and evidence capture.

API reference

Intents, decisions, audits, and exports.

OpenClaw integration

Intercept tool calls and gate execution through Gvner.

Temporal integration

Govern activity execution in deterministic workflows.

HTTP/Webhook adapter

Single contract for non-native runtimes.

LangGraph integration

Gate side-effect nodes through Gvner decisions.

Semantic Kernel integration

Use execution filters with fail-closed governance checks.

Bedrock integration

Govern action group invocation paths with one adapter contract.

Vertex ADK integration

Map intent hooks to Gvner checks before tool execution.

Automation integration

n8n/Make/Zapier fail-closed pre-step guard pattern.

Worker queue integration

Pre-handler guard for Kafka/SQS/Celery style workers.

Exports

Regulator-ready packs and integrity receipts.

Regulator packet

Signed, hashed evidence packets for external reviews.

Security model

Guarantees, controls, and threat coverage.

Full doc tree

Every public control and claim maps to a concrete doc page.

Enterprise trust layer

Public, regulator‑facing materials that make Gvner adoption audit‑ready.

Policy schema (example)

Policies define who can execute, which scopes apply, and the regulatory basis for enforcement.

{ "policy_id": "PB-042", "status": "ACTIVE", "scope": "write_files", "actors": { "allowed": ["first_party_agents"], "denied": ["external_runtimes"] }, "approvals": { "required": 1, "dual_control": true }, "budget": { "max_operations": 1000 }, "regulation": "SOX_RETENTION" }

Evidence model

Decision record

Every decision produces a deterministic record that includes policy basis and evidence hash.

Decision outcome: ALLOW or DENY
Policy basis and regulatory mapping
Immutable audit identifier

Export formats

Evidence can be exported in structured formats for auditors and regulators.

PDF for regulatory submissions
Markdown for internal reviews
JSON for automated audits