Run your software on one stream of signals

Your crashes, user feedback and server errors, next to the outages, releases and advisories of everything you depend on. One stream, one format, so an agent can connect a crash spike to a provider incident or a dependency fix and act on it.

For engineering, SRE and support teams, and the coding and on-call agents that work with them.

Get these events How it works

The events

What your agent listens to

  • Crash in your app

    app.crash_reported

    Planned: first source Overblast feedback SDK (iOS, Android, macOS, Windows, web)

    Crash in Brightdesk Android 4.2.0: NullPointerException in ExportActivity (37 users)

  • Feedback from your app

    app.feedback_received

    Planned: first source Overblast feedback SDK (iOS, Android, macOS, Windows, web)

    Feedback on Brightdesk iOS 4.2: Export to PDF does nothing

  • Error on your servers

    server.error_reported

    Planned: first source Overblast feedback SDK (iOS, Android, macOS, Windows, web)

    api.brightdesk.example: 5xx errors up to 4.1% on POST /v1/exports

  • Service outage started

    service.outage_started

    Ready Provider status pages, worldwide

    Stratapoint Cloud incident: Elevated API error rates in eu-west

  • Service outage resolved

    service.outage_resolved

    Ready Provider status pages, worldwide

    Stratapoint Cloud resolved after 36 min: Elevated API error rates in eu-west

  • Package released

    software.released

    Ready JavaScript package registry (AI and agent SDKs), worldwide; PyPI package registry, worldwide; crates.io package registry, worldwide

    nordvik-sdk 3.0.0 released on PyPI

  • Open-source security advisory

    vulnerability.advisory_published

    Ready Open-source security advisories, worldwide

    GHSA-7x2q-4m9c-hv3p (high): Path traversal in nordvik-uploads

How it works

From event to action

  1. Your apps and servers report crashes, feedback and errors through the Overblast SDK, privately to you.
  2. Provider outages, package releases and security advisories arrive in the same stream.
  3. Your agent groups what belongs together: a crash spike, an open Cloudflare incident, a dependency that just shipped a fix.
  4. It opens the ticket or the pull request, posts in the incident channel, and answers the users who sent feedback.

Wake your agents up. Each event can start Claude Code, Codex or your own script the moment it lands, one by one or in batches: no polling, no cron, no dashboard to check.

Events carry text from public sources. Run agents that act on them isolated, with only the access the task needs. Run them safely

Create your own stream

Tell your agent
npx worldsignals watch --type app.crash_reported,app.feedback_received,server.error_reported,service.outage_started,service.outage_resolved,software.released,vulnerability.advisory_published \
    --agent claude --prompt ./brief.md

# brief.md
When crashes or server errors spike, check for a provider incident or a new release of a dependency that explains it; post the likely cause in #incidents and open a pull request if a dependency fix exists.

Questions

About this use case

Where do crashes, feedback and errors come from?

From the Overblast SDK for iOS, Android, macOS, Windows and the web, and from servers. They are private events: only your account sees them. Sentry, Crashlytics and app store reviews can follow.

What is live today?

Provider outages, PyPI and crates.io releases and GitHub security advisories are collected now. Crash, feedback and error events are planned, with Overblast as their first source.

Do I need Overblast to use the rest?

No. The public DevOps events work on their own; Overblast adds your own apps' signals to the same stream.

More use cases

Give your agents something to notice

Join the waiting list, free during early access. Pick these events and we’ll email you when you’re in.

Get these events See the live stream