Reference library
Answers, specifications, and proof you can inspect.
Use the library to evaluate authorization, human approval, audit integrity, deployment boundaries, and governance operations before a product conversation.
Specification
Guide
Frequently asked questions about Trunnion AI
Direct answers about Trunnion AI, agentic governance, products, human approval, audit evidence, model choice, deployment and commercial scope.
Reviewed 2026-09-20 →Agentic AI for controlled environments
Architecture and diligence guidance for agentic AI handling CUI or operating in regulated, on-premise and disconnected federal environments.
Reviewed 2026-09-20 →Human-in-the-loop approval gates for AI agents
A practical architecture for AI agent approval gates, including risk tiers, parameter binding, expiry, separation of duties and audit evidence.
Reviewed 2026-09-20 →Attribute-based access control for AI agents
How ABAC constrains AI agent actions using subject, resource, operation and environment attributes, with policy examples and evaluation tests.
Reviewed 2026-09-20 →Control mapping
NIST AI RMF mapping for agentic systems
A practical mapping from NIST AI RMF GOVERN, MAP, MEASURE and MANAGE outcomes to architecture, evidence and operating controls for AI agents.
Reviewed 2026-09-20 →OWASP Top 10 for Agentic Applications control mapping
A control-oriented planning map from OWASP's 2026 agentic application risks to enforceable boundaries, tests and evidence.
Reviewed 2026-09-20 →Comparison
How to compare agentic AI governance platforms
A dimension-first comparison framework for AI governance platforms across authorization, audit integrity, approval, deployment and release evidence.
Reviewed 2026-09-20 →Agentic AI versus generative AI
A direct comparison of agentic and generative AI across goals, memory, tool use, human oversight, infrastructure, risk, audit and operating design.
Reviewed 2026-09-20 →Build versus buy agentic AI governance
A decision framework for building a control plane, buying a platform, or combining both without hiding integration and evidence costs.
Reviewed 2026-09-20 →Vertical versus horizontal AI platforms
A practical comparison of vertical and horizontal AI platforms across workflow depth, governance, integration, model choice, deployment and ownership.
Reviewed 2026-09-20 →Cloud versus on-premise versus air-gapped agentic AI
A deployment comparison for governed agents covering dependencies, identity, models, updates, evidence, operations and failure behavior.
Reviewed 2026-09-20 →Documentation
Developer documentation for governed agents
Start with Trunnion's public reference documentation for agent audit records, authorization, approvals, deployment boundaries and verification tools.
Reviewed 2026-09-20 →Agent audit record format reference
Field-by-field documentation for a public AI agent audit record, including identity, authority, approval, outcome, linkage and validation rules.
Reviewed 2026-09-20 →Integration patterns for governed AI agents
How Trunnion evaluates identity, data, model and tool integrations by operation, authority, failure behavior, evidence and release compatibility.
Reviewed 2026-09-20 →Tool
Verify a hash-chained agent audit trail
An interactive, browser-only verifier and test procedure for a public reference hash-chained agent audit format.
Reviewed 2026-09-20 →Agent audit storage and verification calculator
Estimate audit-event volume, storage footprint and verification work for an agentic AI workflow without inventing vendor prices.
Reviewed 2026-09-20 →AI governance readiness assessment
A browser-only assessment for teams preparing governed agentic AI workflows, with concrete gaps and next actions.
Reviewed 2026-09-20 →Commercial guide
Evidence
Field notes
Bring the diligence questions.
We will answer with the release boundary, evidence package, and deployment assumptions in view.