Is it your code, or is it AWS?
OutageDeck reads the official status feeds that 180 cloud and SaaS vendors publish about themselves, about every ten minutes, and turns them into one answer: is it you, or is it them? Paste your package.json on the stack board and get one verdict across everything you depend on. No account, nothing stored.
Or check one vendor now
26 of 180 tracked vendors are reporting problems right now.
150 operational · 4 in maintenance
- AstronomerPartial outage
- AWSPartial outage
- BandwidthPartial outage
- Check PointPartial outage
And 22 more with open problems in the provider directory.
Nothing here is guessed or crowd-reported. Every status is read from the vendor's own official feed and links back to it: AWS Health Dashboard, Cloudflare's status page, OpenAI's status page. How the reading works, including what OutageDeck deliberately does not do, is written down in the methodology.
The providers teams check first
Live status for the most-searched vendors. Every row opens the full sheet: services, regions, 90-day uptime, and incident history.
On the record right now
Current and recently resolved disruptions, kept linkable after the vendor's banner comes down.
- Minor
- Critical
- Minor
- Minor
- MinorPingOne Advanced Services - US East / US West - Logging Systems Degradation
Ping Identity · active
- MinorElevated number of R2 503 errors in Western North America region
Cloudflare · resolved
Told in the channel your team already reads
State changes fan out to Slack, Microsoft Teams, Discord, webhooks, and email, with exactly-once delivery and a retrying ledger behind them. Free covers email alerts for the handful of vendors you cannot afford to miss; paid plans add the team channels and unlimited providers. And any vendor the catalog misses, you add yourself: point OutageDeck at any Statuspage or Instatus status page and it joins your account privately, with the same alerts.
Your coding agent can check your vendors before it debugs your code
Connect the OutageDeck MCP server and Claude Code, Cursor, claude.ai, or any MCP client can ask whether a vendor is down, get one verdict across your whole stack, and read incident timelines. No key needed.
Connect from Claude Code
claude mcp add --transport http outagedeck https://outagedeck.com/api/mcp- get_provider_status
- check_my_stack
- list_active_incidents
- get_incident_details
- get_uptime
- get_outage_report
- search_providers
- list_custom_providers
- add_custom_provider
- update_custom_provider
- remove_custom_provider
- watch_provider
- search
- fetch
A record a vendor cannot quietly rewrite
Incident pages stay linkable after the banner comes down, and uptime is labeled honestly: an event a vendor left open for weeks reads as time it stayed open, not as downtime, because those are different claims. Every vendor is measured on one vocabulary, so "degraded" means the same thing on every page and a postmortem can cite the number without translating it.
See how each category actually behaved · Read the 30-day outage report
Questions teams ask before trusting a status tool
OutageDeck is designed to complement official status pages, not replace them.
- How does OutageDeck decide whether a provider is up or down?
- OutageDeck reads each provider's official status source, the same feed their own status page uses, and translates it into a shared status vocabulary. Every provider page stays easy to compare, and every status links back to the source it came from.
- Does OutageDeck replace the official status page?
- No. Every provider page links back to the official source. OutageDeck adds a consistent public experience, historical context, and provider-to-provider comparability.
- How often is the data refreshed?
- Every provider's official status source is checked about every 10 minutes, around the clock. Pages update moments after a change is detected, and alert subscribers are notified as soon as a change is picked up.
- Can I get notified when a provider goes down?
- Yes. Free accounts get email alerts for up to 5 providers, and paid plans add Slack, Teams, Discord, and webhook alerts across unlimited providers. There is also a versioned JSON API, uptime history, and status badges.
- Can my AI coding agent check vendor status through OutageDeck?
- Yes. OutageDeck runs a remote MCP server that connects Claude Code, Cursor, claude.ai, or any MCP client to live status in one line, with no key required. Once connected, an agent can check a single provider, get one verdict across a whole stack, read incident timelines, and pull independent uptime history before it starts debugging your code.