Feature Requests

The custom measures created in the Dashboard Visuzlizations cannot be added and graphed in the Visializations. Only the predefined metrics could be graphed. Could you please allow to add them as well as the predefined ones?

Pipelines sometimes get data from external sources which have urls. In the console output of pipelines, plain urls will render as a hyperlink because that is core browser behavior since netscape. We have tried links with html, markdown, and mrkdwn (which is what Slack messages use), and these just display the raw syntax. Urls still get hyperlinked by the browser, but In the case of the html and Slack syntax, the browser incorrectly interprets some of the link text as part of the url. (See attached image) What would be ideal is for one or more of these link syntaxes be detected and converted to clickable links by the Harness' console viewer.

Use Case: Our teams maintain independent CI pipelines separate from CD. Additionally, We have a master CI/CD pipeline using chaining. As a product requirement, they want to introduce another layer on top of this—essentially a parent pipeline that invokes the existing CI/CD chaining pipeline.

Have dashboard capabilities that can track the total Stage-level minutes for Builds by infrastructure type and ID. Real world use case: I would like to track how many build minutes each org/project/pipeline is consuming to support a hypothetical charge back model. We have teams using our central K8S clusters to build on Linux, VM Pools for building on Windows and various other combinations. Harness has a usage dashboard for Harness Cloud builds, seeking a similar dashboard for self-hosted infrastructure.

Currently, Harness seems to take the value of a particular field within a values yaml override file that gets referenced last during the deployment. We had some users ask if there's an option or flag that would do a merge of the files then do the deployment. For example, say you had a field called alerts and it's a list. You have two values yaml override files that reference the field. File 1 alerts: db_warning File 2 alerts: memory_warning Is there an option that would merge them (using a feature flag etc.) to: alerts: db_warning memory_warning Currently, only memory_warning would get applied.

This stemmed from our conversation with our Cybersecurity Team asking us to prove how we are restricting access to KMS while using CoSign. Situation: From out understanding, at the moment, if users at the project level decided to write their own signing step to access our account level delegates, they could grab our KMS private key and verify and sign artifacts/evidence. This is not only for KMS but for other features in AWS like AWS Secret. At the moment, there is nothing preventing users at the project level from using our account level delegates to grab our AWS account level services.

Add an ability to ensure an input has a value and can never be "null" when evaluated. This would ensure that pipelines fail if they do not provide a. valid value. Currently if inputs are set as expressions and those expressions are not present or do not contain a value (say a missing pipeline tag is used) they evaluate as null rather than failing. This should be optional to enforce the submission of a value

We need to have mTLS implemented on API calls, not just Delegates. Currently, mTLS is supported for the Delegate only, this is a feature request to extend mTLS support for API calls.

The KMS connector only accepts KMS AR stored in Harness Secret Manager. We need an ability to pull those ARN from AWS secret manager.

We have templates that are used in prod-2 and prod-3 Harness, and have to select the url we use, ( app.harness.io vs app3.harness.io ), in scripts using the account id. I'd like an expression, (for example <+account.base_url>), that will provide the correct base url for whichever Harness instance the template is running in