Erik is the founder of Cloud Posse and creator of Atmos. With over a decade of experience helping teams adopt Terraform at scale, he is passionate about open-source infrastructure tooling and developer experience.
Terraform errors and warnings from uploaded executions now appear alongside CLI exceptions in Issues → Exceptions. Repeated occurrences match a stable issue across repositories, stacks, and components, so you can see where a problem is happening without opening every workflow run.
Stable issue codes and direct links — reference an issue with a workspace-prefixed code such as ACME-67H. Share its dedicated page with your team.
Compact feed — scan severity, status, last seen, event counts, and the number of affected instances. Search by diagnostic text or issue code.
Detailed occurrences — inspect monospaced diagnostics, execution context, affected stacks and components, and links to the GitHub Actions runs. Occurrences without a matching instance are reported separately.
Persistent triage — resolve, ignore, or reopen an issue. A new occurrence after resolution reopens it; ignored issues stay ignored.
Bulk actions — select issues with the checkbox beneath each severity indicator, or select all loaded rows. Resolve, ignore, reopen, set priority, or assign them together. Unsuccessful updates remain selected for retry.
Create a GitHub issue from an exception in one of its affected repositories. Review the generated title and description, then create it with the diagnostic, affected scope, and workflow links included. Atmos Pro saves the GitHub link and shows it in the GitHub Issues view. Exception triage remains independent of the GitHub issue's status.
MCP clients can search issues and occurrences, open an issue by ID, code, or dashboard URL, inspect diagnostic context, count events, explore tag distributions and affected associations, and manage triage and assignment. They can also prepare a GitHub issue draft and create it in an affected repository.
Reads honor workspace permissions and OAuth scopes. Writes require explicit consent and live write permissions, and triage changes are audited. Existing Sentry-compatible client configuration continues to work.
Open Issues to see errors by default. The sidebar separates errors, warnings, and GitHub issues, and links to drifted instances. Filter by repository, stack, component, occurrence time, or issue age. Compact trend bars, first/last-seen commit links, and stack/component composition help locate the affected infrastructure.
Use Ask AI to start an investigation with the selected occurrence and issue context, or Copy as → Markdown to share a diagnostic summary with a stable issue link. Assignment and priority persist across the dashboard and MCP.
Open Account Settings → Notifications to manage multiple personal rules. Search your repository’s inventory and select several stacks or components at once, then edit, enable, disable, or remove each rule independently. Errors are selected by default; warnings are optional. Choose in-app or email delivery for each rule. Wildcards such as payments/* remain an advanced option.
With personal-settings DMs enabled for your Slack installation, message the Atmos Pro bot to manage the same rules. Your account must be linked, and the conversation stays private. Use a channel with the bot for infrastructure questions and other requests.
Slack alerts use each channel's existing repository and stack/component subscriptions. Enable Exception errors or Exception warnings on those subscriptions. New issues and regressions after resolution have distinct alerts, with a stable issue link, the diagnostic, and captured commit/workflow context. Duplicate events, ignored issues, and historical imports do not send alerts.
Email alerts support mail-client one-click unsubscribe and include a preferences link. Unsubscribing removes email delivery from the matching subscriptions while preserving in-app and Slack settings. Sidebar dots show unresolved errors in red and warnings or drift in yellow.
|