Harness has expanded its AI Software Delivery Platform with the launch of the AI Agent Development Lifecycle (Agent DLC), enabling enterprises to build, test, deploy, secure, and govern AI agents using the same software delivery pipelines and controls they already rely on for application development. The announcement addresses one of the biggest challenges in enterprise AI adoption—moving AI agents from experimental pilots into production with governance, security, and operational confidence.
As enterprises increasingly adopt AI agents across engineering, sales, product, and customer support, many struggle to move beyond proof-of-concept projects. According to Gartner, only a small percentage of organizations currently have agentic AI running in production, highlighting the need for enterprise-grade governance throughout the AI development lifecycle.
Harness aims to bridge this gap by extending its AI Software Delivery Platform to support AI agents from development through production while maintaining the same governance, testing, deployment, and compliance controls already established for traditional software.
“When we started Harness, the vision was a safety harness for code,” said Jyoti Bansal, co-founder and CEO of Harness. “Until recently, that meant application code. Today it also means agentic code, written across engineering, product, sales, and support teams alike, each building agents for their own workflows. Everything you've done for software delivery over the last decade — governance, orchestration, security, testing — you can now do for agents in the same platform.”
Unlike conventional software, AI agents operate using large language models that can produce different outcomes even when given identical inputs. This non-deterministic behavior makes testing, debugging, and reproducing issues significantly more complex than traditional software development.
Harness notes that as AI agents gain access to enterprise systems, APIs, and customer data, organizations require stronger governance and visibility to manage security, compliance, and operational risks throughout the deployment process.
The new Agent DLC introduces five capabilities designed to support every stage of AI agent delivery.
Harness AI Evals allows teams to create evaluation datasets, establish scoring functions, and implement automated quality gates that detect performance regressions whenever AI models or agents are updated.
Agent Deployments extends existing deployment pipelines to managed AI agent runtimes, including Amazon Bedrock AgentCore and Google Agent Runtime. Organizations can leverage familiar deployment workflows, including canary releases, approvals, and governance policies.
AI Configs enables organizations to manage prompts and AI model configurations using the same feature flag infrastructure employed for software releases. Teams can compare prompt performance and roll back changes without redeploying applications.
The AI Asset Catalog automatically discovers AI agents, plugins, and skills across enterprise repositories while assigning ownership and improving governance across development environments.
Harness AgentTrace provides detailed visibility into AI agent execution by recording decisions, execution paths, response behavior, and model interactions. Harness is also open-sourcing core AgentTrace components, including harness-sdk and harness-evals, enabling developers to extend tracing capabilities into their own AI applications.
Recognizing that AI agents introduce unique security challenges, Harness also unveiled new security features designed specifically for agentic AI environments.
Shift-left security capabilities include:
Runtime protection capabilities include:
Together, these capabilities provide organizations with a unified audit trail covering AI agents from development through production.
Harness built Agent DLC on top of its Software Delivery Knowledge Graph, which connects data across software delivery pipelines and now extends that intelligence to AI agents.
The platform also builds upon the company's Autonomous Worker Agents introduced in June 2026, allowing enterprises to apply existing deployment approvals, evaluation gates, governance policies, and security controls throughout the AI agent lifecycle without introducing separate operational workflows.
Harness AI Agent Development Lifecycle capabilities are rolling out now for Harness customers.
As enterprises accelerate AI adoption, Harness is positioning Agent DLC as a unified platform that combines software delivery, AI governance, security, deployment, and observability into a single workflow. By extending familiar DevOps practices to AI agents, the company aims to help organizations deploy production-ready AI with greater confidence, visibility, and operational control.
Harness is the AI Software Delivery Platform™ company, enabling engineering teams to build, test, and deliver software faster and more securely. Powered by Harness AI and the Software Delivery Knowledge Graph, the platform brings intelligent automation to every stage of the software delivery lifecycle after code — removing toil and freeing developers from manual, repetitive work. Companies like United Airlines, Morningstar, and Choice Hotels use Harness to accelerate releases by up to 75%, cut cloud costs by 60%, and achieve 10x efficiency across DevOps. Based in San Francisco, Harness is backed by Goldman Sachs, Menlo Ventures, IVP, Unusual Ventures, and Citi Ventures.
