For developers
External HTTP monitoring for the endpoints you ship.
UptimeWire requests your public endpoints from outside, records the status code and response time, confirms failures, and sends outage and recovery events to your tools.
Latest check
- Request
- GET
- Redirects followed
- 1
- Final status
- 200
- Response time
- 184 ms
- Result
- Up
What a check does
A plain GET request, judged by its final status.
Knowing exactly what a check is makes its results easier to trust.
-
It sends a GET request
The request has no custom headers, no body, and no credentials. You cannot choose the method.
-
It follows up to five redirects
The final answer decides. A status from 200 to 399 counts as up. More than five redirects is a failure.
-
It records what happened
Status code, response time, and the reason for a failure are kept for each check. From Starter, each check also shows where the time went.
Pick endpoints that answer without credentials
A 401 or 403 counts as a failure, so a protected route will look down. Monitor a public health route. Production and staging are separate addresses, and each counts as one website.
Look inside the response, a little
A failure can also be raised when the body lacks a text, or contains one. The match is exact and reads the first megabyte. It is not a JSON assertion.
A failed check is repeated about 20 seconds later, and can be repeated from a second location where one is set up, before an incident opens. Confirmation reduces false alarms. It does not remove them.
Events out
Get outage and recovery events where your code can use them.
From Starter, a webhook sends a signed JSON message when an incident opens and when it closes. The signature is HMAC-SHA256 with a secret only you and UptimeWire know.
- One webhook address per account, shared by all your websites
- A message can arrive twice or out of order, so use the id and start time in it
- Webhooks carry outage and recovery only. Certificate, domain, and DNS warnings go to email and Telegram
Other destinations
- Email and Telegram on every plan
- Discord from Starter
- Slack and Microsoft Teams from Business
Note
Webhooks are built and tested against a test receiver. Delivery to a live public address has not been independently verified yet. Send the test message to your own endpoint before you rely on it.
Read-only API and MCP server
Pull your monitoring data into your own tools.
From Business, an API key gives read access to your sites, incidents, and uptime. An MCP server lets an AI assistant read your monitoring data too.
What you can read
- Your websites
GET /api/v1/sites- One website
GET /api/v1/sites/{id}- Its incidents
GET /api/v1/sites/{id}/incidents- Its uptime
GET /api/v1/sites/{id}/uptime
How it works
- Send your key in an Authorization header as a Bearer token
- The API only reads. It cannot add, change, or delete websites
- The MCP server for AI assistants also only reads, and it has its own keys
- There are no official SDKs
Where it fits
One more signal, from the outside.
UptimeWire answers one question well: does the public address respond? Your own tools answer the rest.
What it gives you
- An independent view of availability, not from inside your network
- Status code and response time history for each endpoint
- Incidents with a start, a cause, and a duration
- Certificate, domain, and DNS checks for every address
What it does not do
- Browser tests, scripted journeys, or custom requests
- Logs, traces, error tracking, or server metrics
- CI/CD integration or automatic remediation
- Check a JSON payload against a schema
Plan fit
Free to try, Starter for events, Business for the API.
Try it
Free
1 website, checked every 5 minutes, with email and Telegram alerts.
Events and timing
Starter
10 websites, checked every 3 minutes. Adds webhooks, Discord, and the timing of each check.
API access
Business
25 websites, checked every 1 minute. Adds Slack, Teams, the read-only API, and the MCP server.
Frequently asked questions
Before you start.
Can I send custom headers, a body, or use POST?
No. Every check is a plain GET request. Requests are not customizable.
Can I monitor an endpoint that needs authentication?
Not directly. There are no credentials in the request, so a protected route answers 401 or 403, and that counts as a failure. Monitor a public health route instead.
Does it replace APM, logging, or tracing?
No. It has no logs, traces, error tracking, or server metrics. It adds an outside view of whether your addresses answer.
Is there an SDK or a CI integration?
No. There are no SDKs and no CI/CD integration. You can use signed webhooks and, from Business, the read-only API and an MCP server for AI assistants.
Have the webhooks been tested against a live service?
They are built and tested against a test receiver. Delivery to a live public address has not been independently verified yet, and the integrations pages say so.
Add an outside check to your stack.
Monitor your first endpoint for free. All four checks are included.