Harness AI DevOps Agent

The Harness AI DevOps Agent streamlines your DevOps processes by enabling you to create and edit steps, stages, and pipelines with ease. Leveraging the power of large language models, the agent provides intelligent suggestions, automates repetitive tasks, and now generates and integrates OPA Rego policies to help you meet your compliance standards.

AI Models

The AI DevOps Agents use the following AI models to facilitate your DevOps tasks:

• DevOps Agent: Claude Opus 4.5
• Support Agent: OpenAI GPT-4o
• OPA Agent: OpenAI GPT-4o
• Error Analyzer: OpenAI GPT-4o

Installation and Setup

The Harness AI DevOps Agent is enabled directly within the Harness UI—no separate installation on an external tool or marketplace is required. Follow these steps to activate the agent:

1. Select Account Settings in the left nav.
2. Under General, select Default Settings.
3. Find Harness AI and enable the Harness AI setting.
4. Optionally, enable Allow Overrides. This will allow orgs and projects in the account to optionally override this setting and disable the agent.

note

The DevOps Agent is only available via the Harness UI.

Using Harness AI DevOps Agent: Main Features

Once authenticated, you can leverage the agent’s capabilities to manage your DevOps workflows. The key features include:

Feature	Description
Step Management	Create, edit, and organize individual steps within your pipelines.
Stage Configuration	Modify and configure your stages.
Pipeline Orchestration	Build and update your pipelines across all Harness modules.
Multi-Module Pipeline Creation	Create pipelines for CI, CD, IACM, IDP, SCS, STO, DB DevOps, and Chaos Engineering modules.
Resource Creation	Create and update Services, Environments, Connectors, and Secrets through conversational prompts.
Policy Generation and Integration	Generate and integrate Open Policy Agent (OPA) Rego policies to meet your compliance standards.
Error Analyzer	AI-powered root cause analysis for pipeline failures with automated fix recommendations.
Pipeline Summarizer	Generate natural language summaries of pipelines, executions, and dependent resources.
GitOps Operations	Manage 13 GitOps resource types: query status, sync, create, update applications and ApplicationSets, inspect events, logs, and resource trees, and link clusters to environments.

Step Management

Harness AI DevOps Agent enables you to create new steps or edit existing ones.

• Context-Aware Editing: Edit steps based on your current pipeline configuration.
• Bulk changes: Update numerous steps at once.

Stage Configuration

Configure stages to logically group related steps in your pipelines. The agent supports:

• Creating any stage type within a pipeline.
• Creating any steps within the stage with schema validation.
• Configuring advanced settings like failure strategy, conditional executions, and delegate selectors.

Pipeline Orchestration

Manage your complete pipelines with features including:

• Intelligent Pipeline Creation: Automatically generate pipelines based on project context.
• Seamless Editing: Modify pipelines quickly to adapt to evolving requirements.

Pipeline Generation Demo

For a demo of the Harness DevOps Agent generating pipelines, see the following:

Multi-Module Pipeline Creation

The Harness AI DevOps Agent supports creating pipelines and stages across all Harness modules, enabling teams to build end-to-end workflows that span the entire software delivery lifecycle.

Supported modules and stage types:

Module	Description
CI (Continuous Integration)	Build, test, and scan stages for continuous integration workflows.
CD (Continuous Delivery)	Deployment stages with rollout strategies, approvals, and environment targeting.
IACM (Infrastructure as Code Management)	Infrastructure provisioning pipelines using Terraform and other IaC tools.
IDP (Internal Developer Portal)	Pipelines for developer self-service workflows and automation.
SCS (Software Supply Chain Security)	Supply chain security scanning and attestation stages.
STO (Security Testing Orchestration)	Security testing stages including SAST, DAST, and SCA scans.
DB DevOps	Database schema migration and change management stages.
Chaos Engineering	Chaos experiment stages for resilience testing.
Custom	Custom stages and steps for specialized workflows.

IACM Pipeline Creation

The DevOps Agent can create Infrastructure as Code Management (IACM) pipelines that provision and manage infrastructure using Terraform and other IaC tools. This enables teams to automate infrastructure provisioning directly through conversational prompts.

Example prompts:

• "Create an IACM Pipeline that provisions AWS Infrastructure with the IACM Steps."
• "Build a pipeline with an IACM stage to manage my Terraform workspace."
• "Create a pipeline that runs Terraform plan and apply for my cloud infrastructure."

Cross-Module Pipeline Examples

You can combine stages from multiple modules in a single pipeline:

• "Create a pipeline with a CI stage to build my app, an STO stage to scan for vulnerabilities, and a CD stage to deploy to production."
• "Build a pipeline with a Chaos Engineering stage after deployment to validate resilience."
• "Create a pipeline with an IDP workflow stage to provision a new environment, then deploy my service to it."

Policy Generation and Integration

The Harness AI DevOps Agent also enhances compliance by generating and integrating policies. This capability allows you to:

Policy Feature	Description
Automated OPA Rego Policy Generation	Generate Open Policy Agent (OPA) Rego policies automatically based on your defined compliance requirements.
Seamless Policy Integration	Integrate generated policies into your existing workflows to ensure adherence to compliance standards.

This feature helps maintain robust security and compliance postures by ensuring that your pipelines adhere to industry standards.

Error Analyzer

Harness AI analyzes pipeline failures by correlating recent changes, examining dependencies, identifying historical patterns, and recommending specific fixes. Use this feature to quickly diagnose and resolve pipeline issues without manual debugging.

When a pipeline fails, Harness AI performs:

• Change impact analysis - Identifies recent pipeline modifications that may have caused the failure
• Dependency checks - Verifies the status of external infrastructure and services
• Historical pattern matching - Compares the failure against similar past failures
• Root cause analysis - Determines the specific step and command causing the issue
• Automated recommendations - Suggests prioritized action items with justifications

Analyze a pipeline error

1. Navigate to the failed pipeline execution.
2. Select Analyze Error.
3. Review the analysis results in the Change Impact Correlation panel.

The analysis includes:

• Recent changes with timestamps and authors
• External dependency status
• Historical failure patterns with similarity scores
• Prioritized recommendations
• Impact assessment and risk level

Harness AI provides action items with:

• Priority - High, Medium, or Low urgency
• Action - Specific steps to resolve the issue
• Justification - Explanation of why the action is needed

Each recommendation targets a specific problem, such as:

• Removing intentional failure commands
• Replacing placeholder commands with actual build steps
• Implementing proper error handling

Automatically fix pipeline YAML:

1. In the analysis panel, select Help me fix the pipeline yaml.
2. Review the Pipeline Fix Summary showing:
   • Problem identified
   • Solution applied
   • Before and after YAML comparison
3. Examine the Updated Step YAML in the preview panel.
4. Select Accept to apply the changes to your pipeline.

Use pipeline error analysis to:

• Diagnose failures faster than manual log review
• Identify whether failures stem from code changes or infrastructure issues
• Learn from historical failure patterns across your organization
• Generate fixes for common pipeline configuration errors
• Assess the blast radius of pipeline failures

View audit trail

To see the complete history of pipeline changes:

1. In the analysis panel, select Show Me the Audit trail.
2. Review the chronological list of pipeline modifications.

Use the audit trail to understand the sequence of changes leading to the failure.

Service Creation and Update

Harness AI generates complete service definitions through conversational prompts. Describe your application requirements, and Harness AI creates a fully configured service with manifests, artifacts, and connectors.

Harness AI creates services by:

• Understanding your requirements - Analyzes your conversational input about application type, deployment needs, and infrastructure
• Generating complete YAML - Creates service definitions with all required components
• Configuring manifests - Sets up Kubernetes manifests, Helm charts, or other deployment configurations
• Connecting artifacts - Links to artifact registries with appropriate image references
• Adding metadata - Includes descriptions, tags, and identifiers for organization

Create a service

1. Open the Harness AI assistant in your project.
2. Describe the service you want to create. For example:
   • "Create a Kubernetes service"
   • "Set up a service for my Node.js application"
   • "Generate a service for nginx deployment"
3. Review the generated service definition:
   • Service name and identifier
   • Service type (Kubernetes, ECS, etc.)
   • Manifest configuration
   • Artifact details
   • Connector references
4. Preview the complete YAML in the YAML Preview panel.
5. Select Create to add the service to your project.

Use AI service creation to:

• Quickly onboard new applications without manual YAML writing
• Standardize service configurations across projects
• Generate baseline configurations for customization
• Learn proper service structure through examples
• Migrate applications to Harness faster

Best practices

• Provide specific details - More context leads to better configurations (for example, "Create a Kubernetes service for a Node.js API with 5 replicas")
• Verify connector references - Ensure referenced connectors exist in your project before creating the service
• Update image tags - Replace latest tags with specific versions for production services
• Review resource settings - Adjust CPU and memory limits based on actual application requirements
• Add service variables - Include environment-specific variables after creation
• Use Git sync - Store service definitions in Git for version control

Connector Creation and Update

Harness AI guides you through connector creation with intelligent recommendations for authentication, permissions, and configuration. Describe what you want to connect, and Harness AI helps you set up secure, validated connections.

Harness AI creates connectors by:

• Understanding connection requirements - Analyzes your input about target infrastructure, cloud provider, or service
• Recommending authentication methods - Suggests secure authentication approaches (IAM roles, service accounts, API keys)
• Guiding configuration - Provides step-by-step setup with contextual help
• Validating credentials - Tests connections and identifies permission issues
• Generating complete YAML - Creates connector definitions with all required fields

Connector details

• Name - Human-readable connector name
• Identifier - Unique connector identifier for API and service references
• Type - Connector category (Kubernetes, AWS, Docker Registry, Git, etc.)
• Description - Auto-generated description with creation timestamp

Authentication configuration

• Auth method - Recommended authentication approach:
  • Service account tokens (Kubernetes)
  • IAM roles or access keys (AWS, Azure, GCP)
  • Username/password or personal access tokens (Git providers)
  • API keys (various services)
• Credential references - Links to Harness secrets for secure storage. Harness doesn't accept any actual credentials when configuring a connector. Users will need to update the connector manually to provide secret.
• Permission recommendations - Minimum required permissions for least-privilege access

Environment Creation and Update

Harness AI generates complete environment definitions through conversational prompts. Describe your deployment target, and Harness AI creates a fully configured environment with infrastructure definitions and overrides.

Harness AI creates environments by:

• Understanding deployment context - Analyzes your input about environment purpose, infrastructure, and deployment stage
• Generating complete YAML - Creates environment definitions with infrastructure configurations
• Configuring infrastructure definitions - Sets up clusters, namespaces, regions, or resource groups
• Setting environment type - Classifies as Production or Pre-Production based on context
• Adding metadata - Includes descriptions, tags, and identifiers for organization

Create Environment

1. Open the Harness AI assistant in your project.
2. Describe the environment you want to create. For example:
   • "Create a production Kubernetes environment"
   • "Set up a staging environment in AWS"
   • "Generate a dev environment for my GKE cluster"
3. Review the generated environment definition:
   • Environment name and identifier
   • Environment type (Production/Pre-Production)
   • Infrastructure definitions
   • Connector references
   • Service overrides
4. Preview the complete YAML in the YAML Preview panel.
5. Select Create to add the environment to your project.

Environment details

• Name - Human-readable environment name (for example, production, staging-us-east)
• Identifier - Unique environment identifier for API and pipeline references
• Type - Production or Pre-Production classification
• Description - Auto-generated description with creation timestamp

Use AI environment creation to:

• Quickly set up deployment environments without manual YAML writing
• Replicate environment configurations across projects
• Generate baseline configurations for customization
• Create consistent dev/staging/prod environment structures
• Set up multi-region or multi-cluster deployments faster

Secret Creation and Update

Harness AI helps you create secret configurations by generating the secret object structure with proper naming, scope, and metadata. After AI creates the secret object, you provide the actual secret value manually in the Harness UI for security.

Harness AI creates secret configurations by:

• Understanding secret requirements - Analyzes your input about what credentials or sensitive data you need to store
• Recommending secret types - Suggests text secrets, file secrets, or SSH keys based on your use case
• Generating secret metadata - Creates the secret object with name, description, and scope
• Setting up references - Configures the secret for use in connectors, pipelines, and services
• Guiding value entry - Directs you to provide the actual secret value securely

For security reasons, Harness AI creates the secret object structure but cannot and will not generate actual secret values. After AI creates the secret, you must provide the real credential, token, password, or sensitive data through the Harness UI. This ensures your sensitive information remains secure and under your control.

1. Open the Harness AI assistant in your project or account settings.
2. Describe the secret you need to create. For example:
   • "Create a secret for my Docker Hub password"
   • "Set up a secret for GitHub personal access token"
   • "Add a secret for AWS access key"
   • "Create an SSH key secret for deployment"
3. Review the generated secret configuration:
   • Secret name and identifier
   • Secret type (Text, File, SSH Key)
   • Scope (Project, Organization, Account)
   • Description and tags
4. Preview the configuration in the YAML Preview panel.
5. Select Create to generate the secret object.
6. Immediately provide the secret value in the prompt that appears.
7. Select Save to store the secret securely.

Harness AI generates identifiers following these patterns:

• Secret name - Descriptive name indicating purpose (for example, docker_hub_password, github_pat)
• Identifier - Converts name to lowercase with underscores (for example, docker_hub_password)
• Description - Explains what the secret is for (for example, "Docker Hub authentication password for nginx images")

Use AI secret creation to:

• Quickly set up credential storage with proper naming
• Standardize secret naming across projects
• Create secrets at the appropriate scope level
• Generate metadata for compliance and audit trails
• Organize secrets with consistent tagging

Pipeline Summarizer

Harness AI can generate detailed pipeline summaries by analyzing your pipeline, its executions, and dependent resources. Use these summaries to quickly understand what your pipelines do without manually reviewing each step.

The DevOps Agent analyzes:

• Pipeline configuration and steps
• Recent execution history
• Dependent resources (connectors, secrets, infrastructure)

Harness AI then generates a natural language summary explaining the pipeline's purpose, key steps, and outcomes.

1. Navigate to your pipeline in the Harness platform.
2. Select AI Summarize in the 3 dots menu
3. Review the generated summary.

The summary appears in the pipeline details view and includes:

• Pipeline purpose and goals
• Key stages and steps
• Deployment targets and environments
• Recent execution patterns

Use Cases

Use pipeline summaries to:

• Onboard new team members to existing pipelines
• Document pipeline behavior for compliance
• Quickly understand inherited or legacy pipelines
• Review pipeline changes before approval

GitOps Operations

The Harness AI DevOps Agent provides full operational control over your Harness GitOps environment through natural language. The toolset covers 13 resource types and approximately 25 operations spanning agents, applications, clusters, repositories, ApplicationSets, credentials, events, logs, managed resources, resource actions, dashboards, resource trees, and cluster-environment linking.

Instead of navigating dashboards to check application health or manually initiating syncs, describe what you need in plain language. The DevOps Agent queries your GitOps data, triggers operations, and generates pipeline snippets for GitOps workflows.

Supported resource types

The following table lists all GitOps resource types you can interact with through the DevOps Agent:

Resource type	Description	Available operations
GitOps Agent	An Argo CD agent installed in a Kubernetes cluster. Agents can be scoped at account, org, or project level.	List, Get
Application	An Argo CD application managed by an agent.	List, Get, Create, Update, Sync, Bulk Sync, Refresh, Cancel Operation, Run Resource Action
Cluster	A Kubernetes cluster registered with a GitOps agent.	List, Get
Repository	A Git repository registered with a GitOps agent.	List, Get
ApplicationSet	A template that auto-generates multiple applications from generators. Supports list, git, clusters, matrix, merge, pullRequest, scmProvider, and plugin generator types.	List, Get, Create, Update
Repository Credential	Repository credentials (SSH keys, tokens) for GitOps agents.	List, Get
Application Events	Kubernetes events emitted by a GitOps application (sync events, health changes).	List
Pod Logs	Container logs from pods in a GitOps application's workloads.	Get
Managed Resources	Kubernetes resources (Deployments, Services, ConfigMaps, and so on) tracked by a GitOps application.	List
Resource Actions	Available actions (restart, pause, resume, and so on) for a specific Kubernetes resource within a GitOps application.	List
Dashboard	High-level summary metrics: total apps, healthy/degraded counts, sync status breakdown.	Get
Resource Tree	The full Kubernetes resource tree for an application, showing all resources and parent-child relationships.	Get
Cluster-Environment Link	Links between GitOps clusters and Harness environments.	List, Create, Delete

Application operations

Applications support the broadest set of operations:

Operation	Description
List / Get	List all applications in a project or get details of a specific application.
Create	Create a new GitOps application with a source repo, target cluster, and sync policy.
Update	Change an application's source repo, target branch, destination cluster, or other settings. Linking a service or environment to an application is also handled through update.
Sync	Deploy the latest changes from Git to the cluster for a single application.
Bulk sync	Sync multiple applications in parallel.
Refresh	Force Argo CD to re-check Git and compare with the live cluster state. Supports normal and hard refresh.
Cancel operation	Stop a currently running sync or rollback that is stuck or unwanted.
Run resource action	Perform actions on Kubernetes resources managed by the application (see Kubernetes resource actions).

ApplicationSet operations

ApplicationSets use generators to automatically create multiple applications from a single template. The DevOps Agent supports 8 generator types: list, git, clusters, matrix, merge, pullRequest, scmProvider, and plugin.

Operation	Description
List / Get	List ApplicationSets or get details of a specific ApplicationSet by UUID.
Create	Create an ApplicationSet with any supported generator type.
Update	Modify an ApplicationSet's generators, template, or sync policy.

Example prompts:

• "Create an ApplicationSet using a list generator with dev, staging, and prod environments."
• "Create an ApplicationSet using a git directory generator scanning all folders in my repo."
• "Create a matrix ApplicationSet combining environments (list) and directories (git)."
• "Create a clusters ApplicationSet that deploys to all registered clusters."

Kubernetes resource actions

The DevOps Agent can perform actions directly on Kubernetes resources managed by your GitOps applications. You can first discover what actions are available for a specific resource, then execute them.

• Deployments: restart, pause, resume, scale.
• Argo Rollouts: restart, pause, resume, promote-full, abort, retry, skip-current-step.

Example prompts:

• "What actions can I run on the web Deployment in app my-app?"
• "Restart the web Deployment in app my-app."

Inspect and troubleshoot

The DevOps Agent can retrieve detailed information for troubleshooting:

• Events: View Kubernetes events for an application (sync events, health changes).
• Pod logs: Stream logs from specific containers in an application's workloads.
• Managed resources: List all Kubernetes resources tracked by an application.
• Resource tree: View the full resource hierarchy (Deployment, ReplicaSet, Pod, Service, and so on) with parent-child relationships.
• Dashboard: Get a high-level summary of your GitOps environment, including total app counts, health status, and sync status breakdown.

Example prompts:

• "Show recent events for app my-app."
• "Get the last 100 lines of logs from pod web-abc123 in app my-app."
• "What Kubernetes resources does app my-app manage?"
• "Show the resource tree for app guestbook."
• "How many GitOps apps are healthy vs degraded?"

Query status and health

Ask the DevOps Agent questions about your GitOps environment in natural language. The agent queries across agents, applications, ApplicationSets, clusters, repositories, and the dashboard to answer.

Example prompts:

• "What applications are out of sync? How long have they been out of sync? Which project are the out-of-sync applications in?"
• "What syncs failed in the past 24 hours?"
• "Which applications are unhealthy in the production environment?"
• "List all healthy GitOps agents at account level."
• "List all clusters registered at account level."
• "What clusters are linked to environment prod?"

Trigger operations from chat

You can initiate GitOps operations directly from the AI chat:

• "Sync the app my-app with pruning enabled."
• "Bulk sync apps app1 and app2."
• "Hard refresh all apps on my agent."
• "Cancel the running sync on app my-app."
• "Initiate a sync for all applications that manage non-prod services."
• "Link cluster incluster to environment staging."

Data Storage and Privacy Policies

Harness AI is designed with strict data privacy and security principles. This section explains what data is used, how it is handled, and how it is discarded when you interact with the DevOps Agent or any AI-powered features in Harness. No configuration is necessary to ensure your privacy since Harness applies strict privacy defaults:

• Training is disabled across all AI integrations.
• Data is not persisted or exposed to model providers beyond inference.
• Fallback mechanisms are used only when necessary and are compliant with strict retention policies.

This section discusses these policies in greater detail.

Data Privacy and Subscription Terms

For a full legal breakdown of AI privacy at Harness, see: AI Data Privacy & Subscription Terms 2025.

No Data Use

Harness AI processes real-time user input and relevant context (e.g., pipeline metadata or error logs) to generate responses. This data is not stored, not logged, and never used for training by Harness or its model providers.

Minimum Data Retention

Data discard behavior depends on the underlying AI provider:

Provider	Model	Discard Mechanism	Retention	Used for Training?
Google Vertex AI	Claude Opus 4.5	Immediately purged after inference	0 days	No
OpenAI (fallback only)	GPT-4o	Retained for 30 days, then purged	30 days	No (Harness opts out)

Failover/Fallback Scenarios

In rare scenarios (e.g., outage or capacity limits), Harness AI may fall back to OpenAI APIs:

• OpenAI may retain the data for 30 days.
• Harness does not permit OpenAI to train on this data.
• After 30 days, the data is automatically purged.

No Anonymization/Tokenization Required

No anonymization or tokenization is needed because:

• Data is not stored or used for analytics or training.
• All processing is done in ephemeral memory.
• Discard mechanisms are managed by the underlying model providers.

FAQs

Is this available for SMP customers?

No, Harness AI is not available for SMP.

Who is the AI DevOps Agent available to?

Enterprise Licenses (including Dev360, Service, SI, Users) are entitled to AI DevOps free of charge. Any module that has access to pipelines will be entitled to AI DevOps - not just CI or CD. This includes all Harness modules except for CCM. The scope of the AI DevOps Agent will be restricted to the license you have. For example, a CI-only customer cannot create a CD stage.

Where can I submit feedback?

Please submit feedback by emailing Harness Support or through the UI by clicking Help in the bottom left corner then Give us feedback.

Release Updates

February 18, 2026

The Harness AI DevOps Agent has been upgraded to Claude Opus 4.5, bringing significant improvements across the board:

• Consolidated architecture — Merged 5 sub-agents into a single unified DevOps agent, improving speed, context retention, and pipeline generation accuracy.
• Improved response time and output quality — Faster responses with higher-quality results across all interactions.
• Enhanced template reference capabilities — More accurate and reusable pipeline generation through better template resolution.
• Support for longer and more complex pipelines — Validated with a 50-stage pipeline test, enabling enterprise-scale pipeline generation.
• Stronger pipeline modification fidelity — Higher accuracy and performance when modifying existing pipelines.