title: "Harness Ships Autonomous Worker Agents for DevOps Pipelines" slug: "harness-ships-autonomous-worker-agents-for-devops-pipelines" published: "2026-07-14" beat: "Launches" tags: ["Launches", "Business"] creator: "Agentry Newsroom" editor: "Susanne Sperling, Editor — Human in the Loop" tools: ["Claude (Anthropic)", "Perplexity Sonar"] creativeWorkStatus: "verified" dateReviewed: "2026-07-14" aiActArticle50: "compliant" humanView: "https://agentry.news/harness-ships-autonomous-worker-agents-for-devops-pipelines" agentView: "https://agentry.news/agent/harness-ships-autonomous-worker-agents-for-devops-pipelines"
Harness made its Autonomous Worker Agents and Agent Marketplace generally available on June 30, 2026, letting AI agents run as governed steps inside CI/CD, security, and infrastructure pipelines rathe
Drafted by an AI agent. Verified by Susanne Sperling, Editor — Human in the Loop. AI policy.
Harness, the AI Software Delivery Platform company, announced general availability of its Autonomous Worker Agents and Harness Agent Marketplace on June 30, 2026, marking a shift in how AI operates within software delivery pipelines Morningstar. Rather than functioning as bolt-on scripts or external tools, the agents now run as governed steps inside CI/CD, security, IaCM (Infrastructure-as-Code Management), and testing pipelines, inheriting the same OPA policy enforcement, RBAC, audit trails, sandboxing, and scoped credentials that enterprises already trust for human-run deployments Harness Blog.
Every pipeline step—testing, security, deployment, and remediation—can now run as a reasoning agent instead of a fixed script Let's Data Science. Harness frames this as a fundamental difference: "It's the difference between AI helping write a script and AI running the step." The agents pull context from the Harness Knowledge Graph and act through the Harness MCP Server, giving them access to pipeline state, infrastructure topology, and historical deployment data without exposing raw credentials or secrets to the underlying LLM.
Customers can choose which LLM powers each agent—Anthropic, OpenAI, or Gemini models per pipeline—and all agent actions are logged and auditable for compliance and incident investigation.
The launch includes six prebuilt agents: Autofix, Code Review, and IaCM Remediation among them Harness Blog. The Agent Marketplace allows enterprises to discover, configure, and share additional agents, creating a library of reusable automation without requiring custom scripting for each pipeline variant.
Availability is immediate—both products are generally available to all Harness customers as of June 30, 2026. Existing Harness users with CI/CD, security, or testing pipelines can begin adding agents to pipeline steps without migrating their infrastructure or re-architecting their deployment workflows.
The move addresses a longstanding gap in DevOps automation: most AI agents have operated outside enterprise governance, making them unsuitable for regulated industries or teams with strict audit and change-control requirements. By embedding agents directly into Harness's governed pipeline layer, the company positions autonomous agents as production-grade tooling rather than experimental features.