# Okokumo MCP server

Source: https://glimind.com/server/mcp-registry%2Fcom.okokumo%2Fokokumo
Registry id: `mcp-registry/com.okokumo/okokumo`
Status: live (checked 2026-11-22T16:20:08.047Z)

## Measurements (30 days)

- Uptime: 100%
- Checks: 1020
- Outages (90 days): 0

## Install

```json
{
  "mcpServers": {
    "okokumo-mcp-server": {
      "type": "http",
      "url": "https://app.okokumo.com/mcp"
    }
  }
}
```

Claude Code:

```sh
claude mcp add --transport http okokumo-mcp-server https://app.okokumo.com/mcp
```

## Tools (18)

- `create_check`: Create a check. Needs name; http and tls need url; domain needs url as a bare domain; cron needs cron_expression. Counts against the plan's check limit (Free: 5
- `create_incident`: Declare an incident on the status page. It is a draft (not public) unless published is true: publish only when the user asks. body becomes the first timeline en
- `get_account`: The organization, user and role this API key acts as.
- `get_check`: One check with its current state and settings.
- `get_check_channels`: The alert channel ids a check is restricted to. An empty list means the default: EVERY channel of the organization alerts for this check.
- `get_check_stats`: Latency and uptime of a check over the last day, week or month, as time buckets: points[] of {bucket, avg_latency_ms, uptime_pct, count} where count is the numb
- `get_status_page`: The organization's public status page: its address, whether it is published, and its components (the component ids incidents refer to).
- `list_channels`: The organization's alert channels: id, name, type (email, webhook, slack, discord, telegram) and validated_at (unset = not verified yet, it will not deliver). C
- `list_check_alerts`: The last 20 alert deliveries for a check, newest first: which channel, sent or failed, and why. Use it to answer "was I alerted?".
- `list_check_transitions`: The last 50 state changes of a check (from_state, to_state, reason, reason_code, reason_params, occurred_at), newest first. reason_code is stable and machine-re
- `list_checks`: List every check of the organization with its current state (up, down, unknown), type and settings. heartbeat and cron checks carry the ping_url their job must 
- `list_incidents`: Every incident of the status page, drafts included, newest first, each with its timeline of updates.
- `pause_check`: Pause a check: it stops being monitored and sends NO alerts until resumed. Confirm with the user first. Frees a slot of the plan's check limit.
- `post_incident_update`: Add an entry to an incident's timeline, optionally moving its status (e.g. to resolved). Owners and admins only.
- `resume_check`: Resume a paused check. Counts against the plan's check limit again, so it is refused when the organization is at its limit.
- `set_check_channels`: Restrict which alert channels fire for a check, replacing the current list. [] restores the default: EVERY channel of the organization alerts. There is no way t
- `update_check`: Change a check's settings. Only the fields you pass change; the others keep their current values. Its type cannot change. To pause or resume use pause_check / r
- `update_incident`: Edit an incident's title, status, impact, components, or publish/unpublish/archive it. Only the fields you pass change. To tell readers what is happening, use p

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.
