Skip to main content

Getting Started with Harness AI

Harness AI lets you generate various Harness platform resources like Pipeline, Cloud Cost, CD entities via natural-language prompts.
Use these confirmed sample prompts to quickly create configurations inside Harness.

How it works​

For DevOps, pipeline and platform Automation use cases:

  1. Open Harness AI in your project.
  2. Copy a sample prompt and adjust values (connectors, namespaces, repo names, etc.).
  3. Review the generated YAML, save it, and run your pipeline.

Tip: Be as specific as possible in your prompt—include details like deployment strategy, connector names, or namespaces for best results.

For Cost and Efficiency use Cases:

  1. Open Harness AI in CCM
  2. Copy Sample Prompt and adjust prompt parameters
  3. Review the configuration before creation

Example Prompts for DevOps​

Create Pipelines​

Pipelines define how you build, test, and deploy applications. Harness AI can generate CI pipelines, CD pipelines, or hybrid flows with approvals and rollouts.

Generate a Harness CI→CD pipeline that builds a Docker image and pushes to Dockerhub. Add a CD stage to deploy a Kubernetes service to namespace dev on cluster connector k8s-dev. Include variables for IMAGE_TAG and REPLICA_COUNT, and output only valid YAML

Use case: Combines security scanning into the CI/CD process for Python applications.

Create a pipeline that runs unit tests for python, builds with Dockerfile, scans image with semgrep and deploys to staging using a Rolling strategy. Use GitHub connector rohan-git, Docker connector harness docker hub, and K8s connector k8s-dev.

Use case: Enables GitHub PR validation and continuous delivery on merge events.

Generate a PR-triggered pipeline: on pull_request to main, run a pylint command + python test command, post summary to GitHub Checks, and skip CD. On merge to main, promote image harness:lastest and deploy to dev. Use runtime inputs for tag.

Use case: Standard enterprise promotion pipeline with manual approvals and safer canary rollout.

Create a multi-environment pipeline with a single CI stage and two CD stages: staging → production with a manual approval gate and canary rollout in prod. Parameterize SERVICE_NAME, IMAGE, REPLICAS.

Use case: Parallel testing across multiple Python versions plus Helm-based deployment.

Create a pipeline that uses cache, parallel matrix builds for python versions 2, 3, 3.2 , publishes artifacts to harness:docker, and then deploys a Helm chart with values file values-dev.yaml.

Create Service​

Services define what you deploy (application configuration, runtime, artifacts). AI can generate services for Kubernetes, ECS, Lambda, and more.

Prompts​

Use case: Standardize Kubernetes services using Helm.

Generate a Kubernetes service definition named portal using Helm with chart path /cd/chart/portal. Expose variables: IMAGE, TAG, REPLICA_COUNT. Include artifact reference to <+input> and config files section.

Use case: Build serverless microservices with runtime and configuration baked in.

Create a serverless (AWS Lambda) service named router with runtime nodejs18.x, artifact from <+input>, and environment variables LOG_LEVEL, STAGE.

Use case: ECS-based deployment with resource settings and logging integrations.

Define an ECS service named payments using task def family prod-payment, container harness/payment, CPU 0.25/MEM 2GB, and image harness/payment:latest. Configure logging to CloudWatch group prod_log.

Create Environments​

Environments represent where applications run (dev, staging, prod). Harness AI can scaffold them for you.

Prompt​

Use case: Quickly spin up a non-prod environment to test services.

Generate an Environment named staging (type: PreProduction)

Use case: Quickly spin up a prod environment to deploy services.

Create a Production environment named prod

Create Secret​

Important Security Notes​

  • When creating secrets, Harness will create the Secret object without a value.
  • You must navigate to the created secret in the listing page to provide the value.
  • Harness does not send secret values to the AI model. Sensitive data is stripped before any AI request.

Prompt​

Use case: Create a reusable DockerHub token secret for image pulls.

Create a Secret Text named docker-hub-token. Store in default Secret Manager.

Create Connectors​

Connectors integrate Harness with external systems (Git, Docker, AWS, Vault, etc.). AI can help bootstrap these definitions.

Prompts​

Use case: Automates DockerHub connector creation tied to secrets.

Generate a Docker Registry connector named canary-docker pointing to harness/canary with credentials from secret refs usernameRef: canary and passwordRef: docker-hub-token.

Use case: Simplifies Vault integration for secret management.

Create a HashiCorp Vault connector, called dev-vault and default the rest of the values

Use case: Enables IRSA-based AWS authentication via Harness delegate.

Produce an AWS connector using IRSA with role ARN and delegate selectors.

Reference Pipeline Templates​

Templates allow you to standardize steps and stages across multiple pipelines. Harness AI can reference existing templates by name.

Prompts​

Referencing a Stage Template:

Create a Pipeline with the Deploy to K8s Stage Template.

Referencing a Pipeline Template:

Create a Pipeline with the Golden K8s Pipeline Template

What’s Not Supported​

In DevOps Use Cases​

Currently, Harness AI does not support generating:

  • Infrastructure Definition Creation
  • Template Creation and Update
  • GitOps Application Creation and Update

Example Prompts for Efficiency and Cost​

Create a Perspective​

You can use Harness to Create Cloud Cost Perspectives

Prompts​

Could you create a Cost Perspective rules based on Application costs?

Use Case: Creating a perspective for an application

Could you create a Cost Perspective rules based on Environment costs (dev/staging/prod)?

Use Case: Creating a Service/Component Cost Perspective

Could you create a Cost Perspective rules based on a service/component?

Use Case: Creating a Team/Department cost perspectives to manage team cost

Could you create a Cost Perspective rules based on Team/Department costs?