Multiplayer pipeline editing, Fix with AI everywhere it matters, workspace personalization, stronger run controls, richer integrations, and more transparent operations.
This June ProLog is sourced from Mage Pro release 20260617-0000 and summarizes the biggest product changes from the monthly release window.
The headline is collaboration: teams can now see who is editing a pipeline, avoid block-level collisions, and keep the editor oriented with persistent navigation and restored block visibility controls.
The release also brings Fix with AI closer to failures, gives operators broader trigger and run controls, adds a workspace release notes feed, modernizes Secrets, and expands the integration surface customers rely on every day.
Highlights since the last ProLog
Collaboration gets real-time
Presence, edit locks, autosave warnings, and conflict prevention make pipeline editing feel like a shared workspace instead of a race to save last.
Troubleshooting moves closer to the error
Fix with AI now appears across failed runs, block errors, logs, and editor controls, then seeds Mage Pro AI with the context it needs to help.
The workspace starts where you work
Custom landing pages, persistent editor navigation, better creation flows, and returned block visibility tabs reduce the clicks between idea and pipeline.
Operators get bigger levers
Cancel active runs in bulk, disable active triggers in bulk, and trust schedule state to stay consistent across project YAML and the database.
Integrations keep expanding
New Google Business Profile coverage joins deeper Salesforce Marketing Cloud, Microsoft Ads, Google Ads, and BigQuery improvements.
Admin visibility improves
Trace Audit, published release notes, stronger secret management, and better cloud dev-slot health reporting make change easier to inspect.
What shipped
Multiplayer pipeline editor
Live presence avatars, per-block edit locks, conflict warnings, autosave guardrails, and instant join/leave updates make shared pipeline work safer across tabs and teammates.
Read docs: https://docs.mage.ai/design/data-pipeline-management
AI Sidekick troubleshooting
Fix with AI is now easier to reach from failed runs, block errors, log panels, and pipeline editor actions, with cleaner parsing for multiline Python errors.
Read docs: https://docs.mage.ai/ai/sidekick
Workspace landing page
Teams can choose their initial start screen, including pipelines, monitoring, runs, or triggers, so each browser opens closer to the work that matters.
Read docs: https://docs.mage.ai/guides/developer-ux/workspaces
Pipeline creation and navigation
New Pipeline creation now flows through clearer modals for templates, uploads, and type selection, with random naming suggestions that stay helpful instead of forced.
Read docs: https://docs.mage.ai/getting-started/build-pro-pipeline
Block visibility and editor panels
Block visibility tabs are back, the editor keeps persistent left navigation, and file browsers, audit logs, and secret panels now resize more comfortably.
Read docs: https://docs.mage.ai/guides/developer-ux/code-editor
Run and trigger controls
Operators can cancel all active pipeline runs, bulk-disable active triggers, and rely on cleaner one-off trigger presentation for terminal @once workflows.
Read docs: https://docs.mage.ai/orchestration/pipeline-runs/overview
Secret management
The Secrets dashboard gets modernized sorting, scrolling, clearer create/delete affordances, better empty states, resizable panels, and Doppler-backed interpolation with low-TTL refreshes.
Read docs: https://docs.mage.ai/development/variables/secrets
Data sources and integrations
Google Business Profile, Salesforce Marketing Cloud, Microsoft Ads, Google Ads, and BigQuery all get meaningful improvements across streams, sample data, schemas, metrics, and authentication.
Read docs: https://docs.mage.ai/data-integrations/overview
Release notes feed
Mage Pro image changelogs are now published and rendered on the workspace Home page, keeping release context close to the version your team is running.
Read docs: https://docs.mage.ai/about/releases
Admin trace audit
Admins get a dedicated API activity dashboard with workspace-wide event history, search, filtering, and full detail panels for operational review.
Read docs: https://docs.mage.ai/production/mage/pro
Runtime compatibility
Python 3.11 and 3.12, pandas 2.x, and dbt 1.10.x compatibility all improve, with faster CI and backend/frontend stability work across the stack.
Read docs: https://docs.mage.ai/guides/dbt/overview
Reliability and operations
Process queue sizes reload reliably without requiring a restart.
BigQuery destination jobs respect query API timeouts and fail gracefully when queries hang.
Kubernetes client CPU usage and container memory calculations are more stable in long-running environments.
Scheduler configuration is more fault tolerant, including a fix for cron parsing crashes in status-sorted trigger views.
Block logs stored on S3 and GCS now surface in the Logs panel.
BigQuery naming, metric schema alignment, sample data handling, and service account authentication are more reliable.
OAuth sign-in avoids default role leakage and handles base paths and nested workspaces more cleanly.
Pipeline graph dependency mapping is safer when a block is missing a uuid.
Editor-role users can upload and download files outside the project folder with clearer permission checks.
Pipeline run stats, trigger/schedule actions, UI states, multiline Python comments, connector errors, and test-suite stability all receive fixes.
Source: https://github.com/mage-ai/mage-pro-releases/blob/master/releases/20260617-0000.md