Feature Requests

I want to dynamically generate sections in the IDP workflow based on field values selected in the first title. Currently, the scaffolder doesn't support conditional logic to show or hide entire steps or sections. It would be beneficial to have this feature to enhance the flexibility and efficiency of the workflow.

This is the specific implementation ask for PayPal: AIDA has no awareness of PayPal's internal pipeline patterns, DPE runbooks, OPA rules, or the Raptor/Kraken template structure our CI/CD platform uses. When AIDA gives advice, it gives generic Harness answers that don't apply to our configuration. For SonarQube token failures, AIDA has no knowledge of PayPal's DPE-managed auth pattern. DPE's institutional knowledge in Confluence and DPEIR is completely unavailable to AIDA. We need a customer-owned knowledge base that DPE seeds and maintains: DPE uploads internal runbooks, SOPs, template docs, known-issue guides via a management UI. AIDA indexes PayPal-specific content and prioritizes it over generic Harness answers. OPA policy library awareness — AIDA understands our quality gate rules. Failure → runbook mapping: AIDA surfaces the relevant DPE runbook for a recognized failure pattern. KB management UI editable by DPE without Harness involvement or a support ticket. This must be customer-controlled, not a shared Harness KB. Business value Transforms AIDA from a generic responder into a PayPal-aware assistant. Scales DPE knowledge to 1,000+ developers without proportional support headcount growth. Addresses 54% AIDA uselessness rating.

AIDA currently reads only log text. It has no awareness of infrastructure metrics, no access to customer runbook context, and no memory of historical failure patterns. 54% of users who tried AIDA found it useless for real diagnostic scenarios (Feb 2026 survey). AIDA cannot proactively warn on resource misconfigurations before they cause a failure. We need AIDA to: — Analyze infra metrics alongside logs (not just log text) — Integrate with a customer-provided knowledge base (for PayPal-specific implementation) — Match current failures against historical patterns across builds — Perform root cause chain analysis (infra event → failure cascade → impact) — Provide proactive pre-execution warnings when resource configs are misconfigured This is the general AIDA capability improvement. FR-12 covers the PayPal-specific KB seeding implementation. Business value Converts AIDA from a log text reader into a genuine diagnostic assistant. Addresses 54% uselessness rating from our Feb 2026 survey.

using the filter "modules": ["CI"] in the API https://transunion.harness.io/audit/api/audits/listV2?accountIdentifier is nor giving any CI specific audit events. All the CI activities whatever comes under CI module in harness should be displayed when "modules": ["CI"] " filter is used with teh above API. API doc: https://apidocs.harness.io/audit/getauditeventlist

Project labels needed for perspectives and dashboards for GCP.

Can we have a feature to filter evaluation results of a specific policy?

Request to enable email notification for STAGE and STEP level if we are using Configure Plugin Step (I'm using custom Notifications Stages) Sample: I'm using Configure Plugin Step to get notified for failure and success email notifications, however i need to get notified which stage and which step it failed with reason. Note: I don't want to use Native Notifications Templates.

We use Harness within a restricted environment monitored via proxy. Currently, to avoid access restrictions, our proxy team needs to create rules based on the URLs and behaviors of the accessed pages. For some behaviors, such as monitoring pipeline logs in real-time, the browser opens a data stream, which is not a behavior normally accepted by our proxy team's security rules. I would like to suggest that you evaluate the possibility of creating a standard header for responses coming from Harness, indicating the ID of the account returning the data. This way, we can more securely enable certain types of behavior that are not normally allowed, configuring the proxy to accept responses coming from our accounts.

We use GitOps almost exclusively, and we are interested in adding webhooks from Gitlab to the Argocd server instances to trigger refreshes if possible, as Argocd supports. Is there any internal endpoint we can use for this? I think it is possible with applicationsets as that seems to be a stock applicationset and we can just add a VirtualService to the /api/webhook. Can we do the same for ArgoCD? I understand we can trigger a pipeline, but we were interested in the native ArgoCD functionality for this

As an administrator of Harness CCM I want to limit who can ignore an Asset Governance Rule or a whole Azure subscription so that enterprise-wide governance over resources is maintained. Currently any end user can select to ignore a recommendation and when the dialog opens they are given the opportunity to ignore the rule (ie no more recommendations are generated by the rule) or ignore the subscription (ie all of the resources fulfilling the rule criteria in a subscription will no longer be governed by the rule) or ignore the Resource that this recommendation was about. We only want the last option to be available to the end User. Please remove the options or place them behind a feature flag.