New features and improvements in Atmos Pro.
Protected deployments must be dispatched by Atmos Pro. It rejects unknown run IDs while verifying GitHub commit provenance and enforcing environment branch restrictions on the code being deployed.
Ask Atmos Pro about drift and deployment failures in a private Slack conversation, and follow its progress as it investigates.
Adding or removing a label on a pull request can run the workflows you configure under pull_request.labeled and pull_request.unlabeled.
Declare an optional pr_number input on your plan or apply workflow and Atmos Pro passes the pull request number that triggered the run.
The Repairs page now warns you when a repository has a release workflow but GitHub Issues are turned off, because release status comments cannot be posted there.
Pull request, merge group, and release comments now list each stack under the workflow Atmos Pro dispatched for that event.
Inspect queued and running workflows across your workspace, find competing work, and cancel a batch from one place.
Release events are dated when they were published, and a stack with no workflow for a trigger is shown as skipped instead of failed.
Triage approvals in List, follow a manageable set of changes in Board, or focus on a systematic rollout in Review.
Sessions no longer end when the GitHub token behind them expires, and Enterprise workspaces can cap how long a sign-in lasts.
Get prompt deployment failures and terminal release summaries for the repositories, stacks, and components you follow.
Find Terraform errors, warnings, and CLI exceptions across repositories, see affected instances, and turn diagnostics into GitHub issues.
See created, updated, and destroyed resource counts for each component, with a Total row across your pull request.
Review pull request impact, follow drift and deployment approvals, and ask Atmos Pro for answers and reports in Slack.
Give teammates operational visibility or read-only access to controls and audit records without a Licensed User seat.
The status RSS/Atom feed now only emits an item when there's an actual incident — no more 'All Systems Operational' reposting on every polling cycle.
A new get_workspace_settings MCP tool exposes per-workspace feature toggles (STS, GitHub ChatOps), so a connected agent can explain why a feature is disabled — not just whether your plan includes it.
Top up Atmos AI usage with prepaid credit packs and optional auto-refill.
Run `atmos pro commit` to push generated files back to your PR with a GitHub-verified Atmos Pro App commit that re-triggers CI.
Atmos Pro brokers short-lived, read-only GitHub tokens so CI can pull private Terraform modules and components from your other repos without PATs.
Atmos Pro can now verify drift after successful apply and remediation runs, keeping repository drift status grounded in a fresh detection check.
Bring Atmos Pro's workspace-aware AI into GitHub issues and pull requests: summaries, failure triage, investigations, code-review support, and operational handoffs.
Share reusable Atmos playbooks and durable AI-generated artifacts with your team, available in Atmos AI and through the Atmos Pro MCP across every repository.
Star the repositories and instances you return to most, and they appear right under the Repositories and Instances groups in the sidebar.
Ask a workspace-aware AI assistant about drift, failing workflows, and deployments — without leaving the dashboard.
The Atmos Pro PR comment now includes a plain-English summary of everything the pull request changes across every affected stack and component.
Atmos Pro MCP now gives agents direct access to public docs as searchable, efficient Markdown, making setup, configuration, and troubleshooting faster.
Press ⌘K from anywhere in Atmos Pro to jump to a repo, stack, component, or instance — favorites rank first.
Repositories can now require GitHub-verified commits before Atmos Pro dispatches workflows, and Atmos Pro autocommits are created through GitHub's verified bot signing path.
When a pull request touches a `CODEOWNERS` file, Atmos Pro now validates syntax, every referenced user and team, and duplicate patterns — folded into the existing Atmos Pro check.
Atmos Pro now holds dispatches when a pull request would trigger more changes than your threshold, pauses dispatches when pushes arrive too quickly, and protects GitHub API budget during large fan-outs.
Connect Claude Code, Codex, Cursor, Continue, and Cline to your Atmos Pro workspace through the Model Context Protocol — no API keys, browser-based GitHub login, short-lived revocable tokens.
Atmos Pro now ships a polished light theme alongside dark mode, with a system-aware toggle in the header and user menu.
Atmos Pro now leaves a GitHub PR review — approving when checks pass, requesting changes with the failing stacks and instances when they don't.
The Atmos Pro GitHub App now calls GitHub from a stable, published set of dedicated egress IPs — so customers who enable a default-deny IP allow list on their GitHub organization can inherit our IPs and restrict access to their org to traffic from Atmos Pro.
A GitHub-style calendar view that surfaces failure rate trends at a glance, with color intensity scaled to each day's failure rate.
A dedicated analytics page with DORA metrics, workflow trends, instance health, and change failure rates — all in one place.
Infrastructure instances now have their own page with drift status, bulk drift detection, and direct access from the sidebar.
See every pull request across all your repositories in one view, with Atmos status, swim lane grouping, and search.
Review, approve, reject, or cancel deployment workflows across all repositories from a single dashboard — with bulk actions, full history, and a sidebar indicator so nothing slips through.
Stream workspace audit events to your SIEM or compliance tools in real-time using CloudEvents v1.0 webhooks.
Download official Atmos Pro logos, brand colors, and usage guidelines from a dedicated public page.
Atmos Pro workflows now run inside the official Atmos Docker container, and describe affected works with a single command — no flags required.
Track an entire deployment batch in one place with progress, heatmaps, filters, and run-level visibility.
GitHub run-ID linkage is now native — you can remove the custom run-name line from your workflows.
Atmos Pro now uses GitHub's native OIDC check_run_id to link CLI requests to workflow runs — no run-name encoding required.
We stress tested Atmos Pro with over 2,000 affected components in a single PR, then optimized dispatch, fan-out, and reconciliation based on what we found.
Real-time notifications when your workflows complete, fail, or detect drift — right in the dashboard.
Submit feature requests and bug reports directly from the dashboard with optional screenshot capture.
Open GitHub Issues directly from your infrastructure UI, scoped to the exact stack and component. Assign work, track remediation, and auto-close issues when problems resolve.
Switch between workspaces instantly with the new workspace switcher. Your dashboard context updates immediately with no page reload.
Track infrastructure issues with a dedicated board view. See open issues, what's in progress, and what's been resolved — all scoped to your workspace repositories.
A new board view gives you a kanban-style overview of all infrastructure operations across your repositories, organized by status.
Embeddable status badges that show real-time drift detection status for each component, directly in your GitHub README or any markdown.
Atmos Pro now posts rich, animated PR comments with dependency graphs, status tables, and real-time progress tracking directly in your pull requests.
A searchable audit log that records every action in your workspace — who did what, when, and to which resource.
See which stacks fail consistently so you can focus attention where it matters most and build a more stable system.
YAML is a powerful way to manage infrastructure — it becomes a data model for your architecture. But it's hard to visualize. The Stacks Explorer makes that data model visual, with real-time health status for every component.
A unified dashboard that aggregates metrics and events across all your repositories, giving you a single pane of glass for infrastructure operations.
Live metrics cards show deployment velocity, drift rate, failure trends, and active workflows across your entire workspace — updated in real time.
A redesigned repository view that surfaces key metrics — stack counts, drift status, and recent activity — right on the card, so you can triage without clicking through.
