How to Use Flags in Cubyts?

Introduction

This guide explains how to use Flags in the Cubyts platform to surface risks, understand their impact, and take timely, traceable actions. Flags are AI-generated signals that continuously monitor process, feature, and code health, helping teams convert early risk detection into intentional governance and engineering actions.

Prerequisites

• Access to a Cubyts workspace

• Integrated planning and code tools

• Permission to view and act on flags

Step-by-Step Guide

Step 1: Understand What Flags Are

• Flags are AI-generated signals that evaluate:

  • Process health

  • Feature and planning quality

  • Code and repository health

• They correlate data across planning tools and code repositories to predict risks before they materialize.

• Flags are designed not only to highlight problems, but to drive decisions and corrective actions.
  
  

Step 2: Navigate to the Flags Dashboard

• Open the Flags view in your Cubyts workspace.

• This screen provides a high-level snapshot of all open flags.

• Flags are categorized into:

  • Process flags

  • Feature flags

  • Code flags
    
    

At the top of the page, count-based indicators help you quickly understand where the majority of risks are concentrated.

Step 3: Review Flag Summaries

Each flag in the list shows:

• Flag summary

• Flag type and category

• Impacted artifact (work item, pull request, or branch)

• Severity

• Connected data sources from which the flag was generated

This view acts as a control center for proactive governance.

Step 4: Deep Dive into a Process Flag

1. Select a process flag from the list.

2. The detailed view opens and is structured into three key columns:

   • Reason – Why the flag was raised (for example, planning benchmark misalignment or developer overload)

   • Impact – Potential downstream risks such as sprint spillover, rework, or delivery compromise

   • Supporting Evidence – Traceable data points including related work items, benchmarks, and historical signals

3. Review contextual information at the bottom, such as the sprint the work item belongs to.

This structure ensures decisions are made with full execution context.

Step 5: Take Action on a Process Flag

• You can act on a flag in two ways:

  • Directly from the flags list

  • From within the flag details view

• Example action:

  • Move a work item to another sprint

• When triggered, Cubyts presents a confirmation dialog where you:

  • Select the target sprint

  • Confirm the action

All actions are intentional, auditable, and traceable.

Step 6: Review and Act on Feature Flags

• Feature flags evaluate planning quality across multiple work items using configured guardrails.

• They assess:

  • Story structure

  • Requirement clarity

  • Stakeholder definition

  • Business justification

• The focus is on planning finesse, not just completeness.

Opening a feature flag shows:

• High-level recommendations for engineers

• Actionable guidance aligned with best practices

• Guardrail-level analysis explaining:

  • Which benchmarks were violated

  • Why the story fell short

  • What needs to change to meet expectations

This makes recommendations explainable and defensible.

Step 7: Convert Flag Recommendations into Work

• Users can act on feature flags:
  
  

  • From the flags list, or

  • By accepting recommendations in the flag details view

• When a recommendation is accepted:

  • Cubyts creates a work item in the connected planning tool

  • The work item includes:

    • A clear title indicating it was generated by Cubyts

    • A structured description containing all recommendations

    • A reference back to the originating flag for traceability

• Depending on the planning tool, the item is created as:

  • A sub-task, or

  • A task assigned to the engineer

This ensures AI signals are converted into real, trackable engineering work without manual effort.

How Flags Enable Continuous Governance

• Flags continuously monitor delivery, planning, and code health

• They predict risks early and explain:

  • Why the risk exists

  • What impact it may have

  • What evidence supports it

• They enable direct, auditable actions

• They close the loop by converting insights into execution

Conclusion

Flags in Cubyts transform governance from reactive reporting into continuous, intelligent intervention. By turning AI-generated signals into actionable, traceable work, teams can proactively manage risk, improve execution quality, and ensure governance actions are intentional and defensible.

Video link: https://www.loom.com/share/116875873b2a43b9872aada6290abff1