Product updates

Changelog

What changed in Oobeya, why it matters, and the versions it ships with. Explore the latest updates to engineering analytics, AI visibility, and platform administration.

Claude Code, Codex, and a redesigned Team Scorecard

Connect Claude Code and Codex to Oobeya's AI Impact, monitor assistant usage and costs, and navigate engineering performance through a redesigned Team Scorecard delivery flow. Two new data-health symptoms and severity levels help teams identify gaps and prioritize investigation.

1. Major Updates

1.1. Claude Code: usage, engineering impact, and cost monitoring

New Integration

Claude Code is now available in AI Impact, alongside GitHub Copilot, Cursor, and Codex. Collect assistant metrics through OpenTelemetry across Claude Code plans. In AI Impact, review coding efficiency, PR time to merge, bug count, lead time, and quality indicators alongside assistant activity. Outcome Signals summarizes which metrics moved favorably, stayed stable, or worsened compared with the previous period. These comparisons show co-occurring trends, not proof of causation.

Why it matters

Review Claude Code activity, engineering outcomes, and costs in one workspace, while keeping usage measures, outcome comparisons, and billed spend distinct.

Oobeya AI Impact with Claude Code selected, showing coding efficiency, pull request time to merge, bug count, lead time, quality indicators, and Outcome Signals
Claude Code selected in AI Impact, with engineering metrics and period-over-period outcome signals.
Claude Code usage and activity metrics

In AI Impact → Metrics, track active users, sessions, user prompts, model requests, request errors, developer time, and assistant time. Explore model usage with input, output, cache-read, and cache-creation token breakdowns, plus tool runs and success rates. Team, language, and user views provide context for where Claude Code is used.

Claude Code Metrics tab inside Oobeya AI Impact, showing active users, cost, sessions, prompts, requests, errors, developer and assistant time, commits, pull requests, and activity over time
Claude Code activity summary and usage trends in AI Impact → Metrics.
Claude Code cost monitoring

Monitor Claude Code spending in AI Impact → Cost & Credits, including gross usage value, net billable cost, cost per active user, and usage forecasts. Cost estimates use Claude Code list prices. When a Claude Enterprise API key is configured on the data source, the billed amount is shown as net cost and the difference as a discount.

Oobeya AI Impact Cost and Credits for Claude Code, with October 2026 usage, daily burn rate, forecasts, gross usage value, and net billable cost
Claude Code usage and cost monitoring in AI Impact → Cost & Credits.

1.2. OpenAI Codex: enterprise usage, engineering impact, and token cost tracking

New Integration

The native Codex integration connects your Codex Enterprise workspace to AI Impact using a Codex Admin Key and Workspace ID. Review workspace usage and available cost signals alongside engineering performance, with reporting in AI Impact, Metrics, and Cost & Credits. This complements the existing Codex line-level attribution support in Blamely, which feeds AI Usage analysis.

Why it matters

Bring Codex workspace activity, token consumption, and financial context into engineering reviews using a consistent workspace and reporting period.

Codex usage monitoring and engineering impact

In AI Impact → Metrics, select Codex in the assistant-type filter to review available workspace activity and usage trends. Explore engineering outcomes in AI Impact, including coding efficiency, PR time to merge, quality, and delivery performance. Available measures depend on the connected workspace and collected reports; changes in usage provide context for investigation rather than proof of productivity gains.

Oobeya AI Impact Metrics tab with assistant activity and usage trends over time
Assistant reporting in AI Impact → Metrics. Select Codex using the Type filter to review the connected workspace.
Codex token cost tracking

Open AI Impact → Cost & Credits to review available Codex consumption and financial reporting. Keep the workspace and reporting period consistent when comparing spending with engineering outcomes. Token, credit, and billing measures follow the connected provider’s reporting definitions.

Oobeya AI Impact Cost and Credits reporting interface with consumption, usage forecasts, and financial summary
The Cost & Credits reporting view used to review available consumption and financial signals for the selected source.

1.3. Redesigned Team Scorecard and delivery flow

Major Enhancement

Navigate SDLC stages as a connected delivery flow: Project Management, Development, Code Quality, Security, Testing, Delivery, and Monitoring. The redesigned Team Scorecard brings stage summaries, key metrics, critical signals, and symptoms into a clearer layout. Updated dialogs and lazy loading for analysis and deployment selections make scorecard configuration easier to navigate.

Why it matters

Move from a team-level overview into the stage that needs investigation, with less effort spent locating metrics or configuring analyses.

Team Scorecard overview with coding efficiency trends, technical debt, code quality issues, and a critical unreviewed pull requests signal
Review team performance trends, key metrics, and critical signals in the redesigned Team Scorecard overview.
Team Scorecard delivery pipeline with Development Analytics member work distribution, coding metrics, and a symptoms panel
Navigate the delivery flow into Development Analytics, with member-level work distribution and symptoms alongside the analysis.

2. New Features & Integrations

2.1. New Symptom: S15 — Unreachable Data Sources

Major Feature

A new Warning-level data-health symptom identifies data sources used by a team’s analyses that fail the nightly connection check. It highlights connectivity failures that can leave metrics, scorecards, and other symptoms stale or incomplete.

Why it matters

Identify data gaps caused by unreachable tools and investigate credentials, permissions, URLs, or connectivity before relying on affected reports.

2.2. New Symptom: S16 — Unused Data Sources

Major Feature

A new Info-level data-health symptom identifies configured data source categories that none of a team’s analyses use. It helps teams review gaps in their analysis configuration, including categories whose repositories, boards, or projects have not been selected.

Why it matters

Check whether the team’s scorecard covers the tools relevant to its work, while distinguishing configuration gaps from connection failures.

2.3. Estimated effort in Project Analytics

Major Feature

Project Analytics now supports estimated effort as an additional measurement input.

Why it matters

Include work estimates in project reporting when effort-based measurement is more representative of the team’s work.

2.4. Kanban KPIs in gamification

Enhancement

Gamification now supports Kanban Lead Time, Kanban Cycle Time, Kanban Actual Reaction Time, and Kanban Throughput as automated league KPIs.

Why it matters

Choose performance measures that fit Kanban workflows as well as existing league use cases.

Oobeya league KPI table showing Kanban Lead Time, Kanban Cycle Time, Kanban Actual Reaction Time, and Kanban Throughput with automated scoring
Kanban KPIs available for automated league scoring.

3. Enhancements & Improvements

3.1. Symptom severity: Critical, Warning, and Info

Enhancement

Symptoms now carry explicit Critical, Warning, or Info severity levels. Team Scorecard surfaces critical signals and severity labels so teams can distinguish urgent risks, warnings, and informational findings.

Why it matters

Prioritize investigation without treating every symptom as equally urgent.

Team Scorecard symptoms with Critical, Warning, and Info severity labels
Severity labels distinguish critical review signals, delivery warnings, and informational data-health findings.

3.2. Data source health checks

Enhancement

View the results of nightly connection checks and re-check a data source from the interface.

Why it matters

Investigate connection failures and verify recovery without waiting for the next scheduled check.

Integrations showing the latest connection check time, failed source counts, and Check all now control
Review connection health by integration and see when data sources were last checked.

3.3. Lead time and cycle time in the team tree

Enhancement

The Teams screen now displays Lead Time and Cycle Time metrics in the team tree.

Why it matters

Compare delivery flow while navigating the organization’s team structure.

3.4. Clearer bug severity and team distribution

Enhancement

Sort the Bug Details table by severity. The Bug Distribution by Team chart now separates assigned and unassigned bugs and provides an unassigned filter.

Why it matters

Prioritize higher-severity bugs and identify work that has not been assigned to a team.

3.5. Readable board, project, and assignee information

Enhancement

Board and project names appear in separate columns, board labels are easier to read, and Project Analytics shows assignee email addresses instead of internal IDs.

Why it matters

Recognize the board, project, and assignee behind each record without translating internal identifiers.

Project Analytics list with board and project names in separate columns
Identify each analysis using separate Board and Project columns.

3.6. Developer profiles and account navigation

Enhancement

Developer tags are shown in Individual Scorecard headers. Profile actions are clearer, long account lists can be scrolled in the update dialog, and timezone selectors now support search.

Why it matters

Keep developer context visible and make common profile and settings tasks easier to complete.

3.7. Symptom investigation and repository selection

Enhancement

The root causes tab in Symptoms now supports pagination. Repository filters in AI Usage include tooltips for individual repositories.

Why it matters

Navigate larger investigation results and identify repository selections more easily.

3.8. More useful audit logs

Enhancement

Audit transaction logs now provide improved information about platform activity.

Why it matters

Review administrative actions with clearer audit context.

Production delivery metrics, AI costs, and individual leagues

Separate production delivery signals, track engineering trends and Copilot costs, and manage individual leagues with new administrative controls.

1. New Features

1.1. Production-Aware DORA Metrics

Major Feature

Designate production pipelines and measure Deployment Frequency and Lead Time separately from non-production pipelines.

Why it matters

Get a more accurate view of production delivery performance by distinguishing production releases from other deployments.

Oobeya production DORA metrics with Production and Development tabs, lead time breakdown, and deployment frequency chart
Production delivery metrics with a separate view for development pipelines.

1.2. Team Scorecard Trend Reports

Major Feature

Track how engineering performance metrics evolve over time with new trend reports in Team Scorecard.

Why it matters

Identify improvements, detect negative trends, and evaluate the impact of engineering initiatives over time.

1.3. AI Cost & Credits

Major Feature

Monitor GitHub Copilot usage costs and AI credit consumption through a dedicated AI Cost & Credits page.

Why it matters

Understand AI spending and consumption patterns to inform investment and cost management decisions.

GitHub Copilot Cost and Credits dashboard showing credit consumption, total cost, usage forecasts, and risk insights
Copilot credit consumption, costs, forecasts, and usage risks in one view.

1.4. Individual Leagues

Major Feature

Gamification now supports individual leagues, with applicable performance KPIs and tag-based participant filtering.

Why it matters

Recognize individual contributions, encourage developer engagement, and run more targeted gamification initiatives.

Oobeya league selection with Individual Leagues and Team Leagues based on engineering KPIs
Choose individual or team leagues based on engineering KPIs.

2. Enhancements & Improvements

2.1. Refreshed Team Scorecard & Project Analytics

Major Enhancement

Explore redesigned analytics views with standardized widget layouts, clearer KPI summaries, and more intuitive navigation.

Why it matters

Compare metrics more easily and reduce the effort required to interpret engineering reports.

2.2. Configurable Metric Calculations

Enhancement

Administrators can now choose weighted-average calculations for aggregated metrics across work items. Simple averages remain the default.

Why it matters

Choose the aggregation method that best represents your reporting and measurement needs.

2.3. Developer Tag Management

Enhancement

Organize developers by seniority, such as Senior, Mid, or Junior, and specialization, such as mobile, full-stack, or backend. Create and maintain these tags in a centralized catalog, then assign them to multiple developers at once or manage them on individual profiles.

Why it matters

Use tags to filter league participants by experience level or specialization, making it easier to organize targeted individual leagues and maintain developer groups as your organization changes.

2.4. League-Based Access Controls

Enhancement

Control league visibility and management through dedicated permissions for administrators, referees, and participants.

Why it matters

Manage leagues with clearer responsibilities and controlled, role-based access.

2.5. GitHub Bulk Import via Excel

Enhancement

Manage GitHub add-on bulk operations using Excel-based templates, including template generation, download, and import for analysis.

Why it matters

Accelerate bulk configuration, reduce repetitive manual work, and simplify GitHub onboarding for large engineering organizations.

3. New Integrations

3.1. Zephyr Scale Cloud Integration

New Integration

Connect Zephyr Scale Cloud to extend Oobeya's test analytics capabilities.

Why it matters

Bring additional testing data into engineering reporting and improve visibility into software testing activities.

Need help with an update?

Read the documentation Contact us