Guardrails
Rate limits, circuit breakers, and budgets that halt unsafe or runaway agent behavior before it hits production or your invoice
Exemplar
How this harness capability fits the Exemplar platform—governed agent operations, not a standalone prompt playground.
Why Exemplar
A well-worded prompt cannot replace enforcement: agents will loop, over-call APIs, and bypass intent when pressure is high.
Guardrails are the harness layer—rate limits, breakers, and budgets applied uniformly across console, webhooks, and MCP.
What Exemplar delivers
Automated circuit breakers that kill runaway sessions before massive token or API cost accrues.
Tier-aware budgets and approval gates so tier-0 services cannot be changed by an agent without human co-sign.
How teams use it
Set soft budgets for exploration and hard caps for autonomous runs; configure breakers on high-risk tool chains.
Blocked actions return structured reasons (which gate failed) so agents retry with corrected context—not opaque refusals.
Capability checklist
Developer guide
Official documentation on docs.exemplar.dev for this capability.
Open developer guide (opens in a new tab)Contact sales
Harness Platform is scoped per deployment. Talk to us about this feature.
From the blog
Related posts on exemplar.dev.
- Agents, context, and guardrails on a unified platform
From code completion to production actions; Context Lake, catalog, governance, and Agentic Assistant/MCP for safe automation.
- AI Agent Governance: How to Control AI Agents Running in Production
AI agent governance is the set of policies, controls, and audit mechanisms that determine what AI agents can do, when they need human approval, and how their actions are logged. The five pillars, how governance differs from the harness, and why it matters for compliance.
- The Harness Engineering Checklist
15 things to put in place before trusting AI-generated code in production — organised by phase: foundation, enforcement, task design, and maintenance. The checklist most teams wish they had before they started.
- Agent loops, tokenomics, and the harness
Why the model is no longer the product: the loop turns intelligence into work, the harness governs it, and tokenomics (token value per watt per user) decides whether it pays. Field examples from Perplexity CEO Aravind Srinivas on 20VC.