On this page
A deploy fails. Before you change anything, you want to know whether a vendor your application depends on is reporting trouble.
Statusfield’s read-only REST API and hosted MCP server let you ask that question from a script or a compatible AI assistant. Search the catalog, read status for vendors your workspace monitors, and list monitors reporting problems.
The result gives you somewhere to investigate. A vendor incident alone cannot tell you whether your code caused the failure.
What is available
Both interfaces use one workspace API key and are included on Starter, Pro and Team, including eligible trials. See pricing for current limits.
- REST API at
/api/v2/*: search catalog metadata, read monitored vendor status, and list your workspace’s monitors. - Hosted MCP server at
/api/mcp: three read-only tools for catalog search, monitored service status, and your monitors reporting problems.
Vendor status comes from official provider reports. These interfaces read the records Statusfield has collected; calling them does not run a fresh website check.
REST quickstart
Create a key in Integrations → API Keys. Copy it when it is created; it is shown only once.
Read workspace monitors with non-operational states:
curl -H "Authorization: Bearer sf_your_key" \
"https://statusfield.com/api/v2/monitors?filter=outages"| Endpoint | Returns |
|---|---|
GET /api/v2/monitors | Your active monitors and their recorded status; add ?filter=outages to filter for non-operational states |
GET /api/v2/services?search=…&limit=… | Catalog metadata matching a name |
GET /api/v2/services/{slug} | Recorded status and components for a service your workspace actively monitors |
GET /api/v2/outages | Non-operational service-wide records for catalog services your workspace actively monitors |
Catalog search does not grant status access. A service-status lookup without an active workspace monitor returns 403 service_not_monitored.
Rate limits and monthly quotas depend on your workspace plan. Read the X-RateLimit-* and X-Quota-* response headers and follow the API reference for pagination, errors and freshness fields.
MCP quickstart
Connect Claude Code using your workspace key:
claude mcp add --transport http statusfield https://statusfield.com/api/mcp \
--header "Authorization: Bearer sf_your_key"This uses Claude Code’s remote HTTP connection with a Bearer header. Keep the key out of prompts and shared files. Other clients need support for remote Streamable HTTP and an Authorization header; see the MCP setup guide.
| Tool | What it does |
|---|---|
search_services | Find catalog metadata by service name |
get_service_status | Read recorded status for a service your workspace monitors |
list_my_down_services | List your active monitors with non-operational states |
Example: investigate a failing deploy
The following is an illustrative conversation, not a live incident or a guaranteed assistant response:
You: Is anything I monitor reporting problems that could affect this deploy?
Assistant: The tool returned GitHub Actions with degraded performance. That is a dependency to investigate. Check the vendor’s incident details alongside the deploy logs before deciding what caused the failure.
If Stripe is not monitored in your workspace, the assistant can find it in catalog search, but cannot retrieve its status through get_service_status. Add the monitor in Statusfield first. These tools cannot add it for you.
What the result tells you
The API and MCP server make monitoring records available inside your existing workflow. They do not establish that your entire stack is healthy or that a deployment is safe.
An empty problem list means the query returned no matching records. Check that the dependencies you care about are monitored. MCP status text does not include observation timestamps; use the REST API or service detail page to check freshness.
For failures specific to your own application, add public HTTPS uptime checks. For scheduled work that can stop without an error reaching you, explore heartbeat monitoring and its access requirements.
Common questions
Can an assistant change my monitors or account? No. The three MCP tools and REST API v2 are read-only. They cannot create monitors, change settings or send notifications.
Does this include every vendor’s status without setup?
No. You can search the catalog, but status access requires active monitors in the API key’s workspace. For all your active monitors, including operational ones, use GET /api/v2/monitors.
Does it detect a vendor outage before the vendor reports it? Vendor monitoring follows official reports. REST and MCP let you read those records programmatically; they do not change the source or detection method.
Start with one monitored dependency
Choose a plan with API and MCP access, add a vendor your team relies on, and create a workspace key in Integrations → API Keys. Try a status lookup and confirm that the returned scope matches what you intended to monitor.
Eligible first-time customers can use API and MCP access during a trial of an included plan. Continued access requires a paid subscription.
Questions about your setup? Contact support.
Keep your vendors in view.
Choose from 7,000+ services, follow the components you use, and configure alerts for your team.
Try any paid plan for 30 days, no credit card required. Free trial for first-time customers.