# Today's Outage > Live tracker of what is broken today: outages from 40+ official status pages (Apple, AWS, Azure, Google Cloud, Google Workspace, Microsoft 365, Cloudflare, OpenAI, GitHub, Slack, Discord, Zoom, banks/fintech and more) merged with outage news headlines, ranked into one "today's outage". Free to use, no key. Data changes about once a minute; please do not poll faster than every 60 seconds. Bursts above 20 requests per 10 seconds per IP to `/api/*` or `/mcp` get HTTP 429 for 10 seconds. "Today" is the current US Eastern (America/New_York) calendar day. Official status pages are polled every 2 minutes, news every 5 minutes; headlines are grouped by organization with AI (Cloudflare Workers AI), with a keyword fallback. ## Data - [Markdown summary](https://todaysoutage.com/today.md): today's outage, all outages ranked, open status-page incidents, status board, latest headlines, last 7 days. Also returned by `GET /` with `Accept: text/markdown`. - [JSON snapshot](https://todaysoutage.com/api/status.json): the same data, structured. Key fields: `headline` (today's top outage or null), `outages[]` (ranked; `status` is ongoing | resolved | unclear; `score` is relative loudness; `joke` is the "unhinged mode" one-liner, null where jokes are off-limits), `ongoing[]` (open official incidents), `services[]` (status board: up | degraded | down | unknown; `checked_at` is the last successful poll of that status page, `fresh` is false when the page stopped answering and the status is stale), `wire[]` (headlines, last 24 h), `week[]` (last 7 days, oldest first; the last item is today), `days_since_major`, `generated_at` (epoch ms). - [RSS feed](https://todaysoutage.com/feed.xml) - [OpenAPI description](https://todaysoutage.com/openapi.json) ## Pages - `https://todaysoutage.com/is--down`: live "Is X down?" answer (Yes / Maybe / Not anymore / No / Probably not) for 80+ services, e.g. [Is AWS down?](https://todaysoutage.com/is-aws-down), [Is ChatGPT down?](https://todaysoutage.com/is-chatgpt-down), [Is Gmail down?](https://todaysoutage.com/is-gmail-down). Full list in the [sitemap](https://todaysoutage.com/sitemap.xml). - `https://todaysoutage.com/day/YYYY-MM-DD`: archive of each past day's outages. ## MCP server Endpoint: `https://todaysoutage.com/mcp` (Streamable HTTP, stateless, JSON responses, no auth). Tools: - `get_todays_outage`: the top outage today plus every other outage today, ranked. - `check_service` `{ "name": "AWS" }`: is this service or company having problems right now? Returns `verdict` (yes | maybe | mostly | fixed | no | quiet; `mostly` = working, only minor issues open), a one-line `answer`, evidence, and the matching `/is--down` page. - `list_ongoing_incidents` `{ "category"?: "cloud", "provider"?: "GitHub" }`: open incidents on official status pages. - `get_outage_news` `{ "limit"?: 10 }`: latest outage headlines. Example client config: `{ "mcpServers": { "todaysoutage": { "type": "http", "url": "https://todaysoutage.com/mcp" } } }` ## Caveats - Status pages often lag real outages by 30+ minutes; news headlines can be wrong or speculative. - `outages[].status` from news alone is a best guess (`unclear` when unknown). - Not affiliated with any company listed.