Frequently asked questions.

Nomotic is the Agent Management Platform (AMP) for AI agents — the only platform that hosts your agents and governs them. It gives every AI agent in your organization a verified identity, manages its full lifecycle, evaluates every action across 20 governance dimensions before it executes, and maintains an immutable audit trail of every decision.

Think of it as the governance layer that decides not just what agents can do (permissions), but what they should do — given their behavioral history, trust level, organizational context, and policy. We call this the Should Layer™.

Guardrails filter LLM outputs after they're generated. They check whether a response contains harmful content, PII, or off-topic material. That's valuable, but it's not governance.

Nomotic governs agent behavior before, during, and after execution. It evaluates whether an agent should take a specific action based on 20 dimensions of risk, the agent's behavioral history, its trust trajectory, and organizational context. Guardrails can't say "this agent's behavior has been drifting for the last 200 actions — tighten its authority." Nomotic can.

Claude Managed Agents and Deep Agents Deploy are execution platforms — they handle deploying and running agents. Nomotic is a governance platform — it handles what those agents are allowed to do once they're running.

Nomotic works with both. Whether your agents run on Claude's infrastructure, LangChain's, or your own, Nomotic governs the decisions they make. You keep your choice of execution platform; Nomotic ensures every agent decision is evaluated, audited, and accountable.

Traditional access control answers "is this agent allowed to do this?" Behavioral governance answers "should this agent do this, right now, given everything we know about how it has been behaving?"

An agent might be technically authorized to send emails — but if it has been sending 10x its normal volume over the last hour, that's a behavioral signal that warrants review. Nomotic tracks behavioral patterns over time and uses them to inform every governance decision.

Every governance evaluation is an optimization problem: minimize the expected cost of the decision given the available information. Blocking a safe action has a cost (lost productivity). Allowing a dangerous action has a cost (potential harm, liability, compliance violation). Nomotic's evaluation framework weighs these costs across all 20 dimensions.

This extends to escalation. Nomotic escalates to human review when the expected value of human input exceeds the cost of the delay — not always, not never, but when the math says it matters. We call this VOI-based escalation (Value of Information).

The Nomotic AMP is organized around six pillars:

• Identity — Every agent gets a cryptographically signed certificate before it takes any action. Anonymous agents are ungovernable.
• Lifecycle — Manage agents from provisioning through deployment, monitoring, and retirement across sandbox, staging, and production zones.
• Governance — 20-dimensional behavioral evaluation with interrupt authority, escalation paths, and a hash-chained audit trail.
• Context — Inject your organization's brand guidelines, compliance rules, domain knowledge, and behavioral boundaries into every governance decision.
• Orchestration — Schedule agents, configure triggers, manage multi-agent workflows, and control execution priority.
• Intelligence — Behavioral analytics, drift detection, governance recommendations, and the Behavioral Context Window.

An agent certificate is a cryptographically signed identity record issued when you register an agent with Nomotic. It establishes who the agent is, what it's authorized to do, and binds its identity to its governance configuration. Certificates use Ed25519 signatures and are hash-bound to the governance policies in effect at issuance.

Just as SSL certificates authenticate websites, agent certificates authenticate AI agents — providing verifiable proof of identity that persists across every action the agent takes.

The Behavioral Context Window is Nomotic's approach to agent memory in governance. Rather than evaluating each action in isolation, Nomotic looks at the last N evaluations for an agent (configurable: 10, 25, 50, or 100) to understand behavioral baselines, detect anomalies, and identify trends.

If an agent's allow rate drops 20% week-over-week, or its average confidence score is declining, or it suddenly starts taking action types it has never taken before — the Behavioral Context Window surfaces these patterns and feeds them into governance recommendations. Your audit trail isn't just a compliance record. It's the behavioral memory that informs every future governance decision.

Agents change over time. An agent that was processing 70% read actions might gradually shift to 50% write actions. That's behavioral drift. Nomotic detects drift across action types, targets, temporal patterns, governance outcomes, and semantic meaning.

Critically, Nomotic also detects human drift. If human reviewers start rubber-stamping approvals — rising approval rates, declining reversal rates, increasing consecutive approvals without challenge — that's a governance gap. Nomotic monitors both sides of the loop.

Radar is Nomotic's agent discovery tool. It scans your organization's network and infrastructure for AI agents that are operating without governance — agents that haven't been registered, certified, or brought under policy. Shadow AI is one of the fastest-growing enterprise security risks. Radar finds it before it finds you.

Your agent certificates, audit trail, and governance history remain intact — your data is yours. If you downgrade to the Free plan, agents above the free tier limit become inactive (not deleted) until you upgrade again or reduce your active agent count. You can export your full audit trail at any time from Settings.

Average evaluation time is under 3ms across all 20 dimensions. The three-tier evaluation cascade means simple cases — clear pass or fail on hard boundaries — resolve in microseconds at Tier 1. Only ambiguous cases escalate to weighted scoring at Tier 2 or full deliberation at Tier 3. Most actions never reach Tier 3.

Every agent action is evaluated across 20 behavioral dimensions that cover scope compliance, authority verification, resource boundaries, behavioral consistency, cascading impact, stakeholder impact, incident detection, isolation integrity, temporal compliance, precedent alignment, transparency requirements, human override, ethical alignment, jurisdictional compliance, and six additional dimensions in the advanced evaluation layer.

Several dimensions carry veto authority — they can block an action regardless of how well it scores elsewhere. You can't override a scope violation with strong ethics scores.

Take any agent's actual behavioral history and replay it against a different governance configuration. "What if we had applied stricter thresholds during last week's incident?" "How many additional denials would the new policy produce over the last 30 days?"

Counterfactual replay answers the "what would have happened if" question that every regulator, auditor, and risk officer actually wants answered — without needing to run experiments on live agents.

Nomotic maintains governance authority throughout execution — not just at the point of decision. If an agent's behavior drifts during a long-running task, Nomotic can signal an interrupt at safe checkpoints. The agent reaches a safe state before stopping, rather than being forcibly terminated mid-transaction, which could leave partial state or corrupted data.

All platform data — agent certificates, governance evaluations, audit trail, context profiles — is stored on Nomotic's infrastructure in the United States. Enterprise customers can request EU data residency or air-gap deployment for complete data isolation on their own infrastructure.

Agent execution data (what your agents actually do) stays within your environment. Nomotic only sees the action metadata submitted for governance evaluation, not the underlying content.

Nomotic is framework-agnostic. It includes integration guides and adapters for LangChain, LangGraph, CrewAI, AutoGen, OpenAI SDK, Anthropic Claude, MCP, Claude Managed Agents, Deep Agents Deploy, and HTTP Proxy for any other framework. If your agent can make an API call, it can be governed by Nomotic.

Basic governance integration takes about 2 minutes — register your agent, get your API key, and add the governance call to your agent's execution path. The platform walks you through it step by step.

For a full production deployment with custom context profiles, organizational policies, and team governance workflows, plan for a few hours of configuration. Most teams are fully operational within a day.

Yes. Nomotic exports to OpenTelemetry for metrics and traces, and supports CEF, Syslog, and JSONL formats for integration with Splunk, Datadog, Elastic, and other platforms. The audit trail is also queryable via API for custom integrations. Webhooks are available for real-time event delivery to any endpoint.

Yes. Claude Managed Agents handles execution infrastructure. Nomotic handles governance. The two are complementary — Claude's platform decides how your agent runs; Nomotic decides whether each action it takes should be allowed. Your agent gets cryptographic identity, behavioral evaluation, and an immutable audit trail regardless of where it's hosted.

Yes. The audit trail is hash-chained using SHA-256. Each record contains the hash of the previous record, creating a cryptographic chain where any modification, deletion, insertion, or reordering is immediately detectable. Governance seals provide signed, single-use authorization artifacts that serve as cryptographic proof that governance evaluated and approved each action.

This architecture meets the evidence standards for EU AI Act Article 12, SOC 2, and HIPAA audit requirements.

Article 9 (risk management): Continuous evaluation across 20 dimensions with behavioral drift detection. Article 12 (record-keeping): Hash-chained audit trail with complete governance provenance. Article 13 (transparency): Reasoning artifacts and governance scorecards. Article 14 (human oversight): Approval queues, bidirectional drift detection, and interrupt authority. EU AI Act reporting is available on the Business and Enterprise plans.

Your governance data — evaluations, audit trail, certificates, context profiles — is scoped to your organization and is not accessible to other customers. Nomotic's internal team can access account data for support purposes only, subject to our privacy policy. Enterprise customers can request contractual data access restrictions and audit rights.

Yes. Audit trail export is available on Team and above plans in CSV, JSON, and CEF formats. You can export by date range, agent, verdict type, or full history. The export includes the full hash chain so the integrity of exported records can be independently verified. Enterprise customers can configure continuous export to their own storage.

The Free plan includes up to 5 managed agents, 10,000 governance evaluations per month, full 20-dimension behavioral evaluation, agent certificates, lifecycle management, cloud audit trail, and community support. There's no time limit — the Free plan stays free as long as you're within the limits.

It's designed to be genuinely useful for individuals and small projects, not a stripped-down trial that forces an upgrade on day one.

Free: 5 agents, 10K evaluations/mo, 1 user. Team ($149/mo): 25 agents, 100K evaluations/mo, 5 users — adds audit export, custom archetypes, context profiles, drift alerts, Radar, and email support. Business ($499/mo): 250 agents, 1M evaluations/mo, 20 users — adds Arbiter advisory layer, VOI escalation, multi-agent coordination, governance scorecard, and 99.9% SLA. Enterprise: Unlimited everything, custom pricing, dedicated support.

You'll receive an alert when you reach 80% of your monthly evaluation limit. If you reach the limit before the month ends, further governance API calls will return a plan limit error until the next billing cycle or you upgrade. Your agents don't stop running — governance evaluation is blocked, not agent execution. We recommend upgrading before hitting the limit for production workloads.

Billing is monthly, charged on your subscription anniversary date. Upgrades take effect immediately with prorated charges for the remainder of the billing period. Downgrades take effect at the end of the current billing period — you keep full access until then. All billing is handled securely through Stripe. You can manage your plan, update payment methods, and download invoices from Settings → Billing.

Yes. Enterprise plans include unlimited agents and evaluations, custom governance dimensions, EU AI Act and FINRA/Federal Reserve compliance presets, air-gap and on-premise deployment, dedicated support with SLA guarantees, custom contracts, and volume pricing. Contact us to discuss your requirements.

Every Nomotic account has a unique referral link in Settings → Refer & Earn. When someone signs up through your link and activates their first agent, they receive one free month on the Team plan. You earn rewards based on how many successful referrals you make: 1 activation earns +2 bonus agents on your plan, 2 activations earns +4 agents, and 3 activations earns one free month of Team.