stillonline-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@stillonline-mcp list my projects"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
stillonline-mcp
MCP server for StillOnline — hosted uptime monitoring and status pages for indie SaaS and production AI agents.
Use it from Cursor, Claude Code, Windsurf, Codex, Antigravity, or any MCP-compatible client. Wraps StillOnline REST API v1 (Pro/Ultimate).
Product | |
MCP docs | |
npm | |
Official MCP Registry |
|
Smithery | |
Demo status page |
Quick setup
Run in PowerShell or Terminal (not in the AI chat):
npx -y stillonline-mcpSame as npx -y stillonline-mcp setup — the wizard starts when you run the package in a terminal. Editors launch the MCP server over stdio separately.
The wizard will:
Confirm the package from npm
Arrow keys + Enter — Cursor, Antigravity, Codex, Claude Code, Windsurf, or Other agent
Cursor / Antigravity / Windsurf — writes MCP JSON config
Codex / Claude Code —
codex mcp add/claude mcp add+ snippetVerify the API key against StillOnline
Empty or invalid key → error and prompt again.
PowerShell paste: hidden input often blocks Ctrl+V — use right-click, Shift+Insert, or:
npx -y stillonline-mcp setup --visible-key --ruThen reload MCP in your editor (e.g. Cursor Settings → MCP → Refresh).
Options
npx -y stillonline-mcp setup --client cursor --ru
npx -y stillonline-mcp setup --dry-run
npx -y stillonline-mcp setup --yes # overwrite existing stillonline blockRelated MCP server: DevHelm MCP Server
Requirements
StillOnline account with Pro or Ultimate
API key from Settings → API (
sk_live_…, shown once)
Security
Do not paste your API key into the AI chat. Use setup or MCP env on your machine only.
Manual MCP config
{
"mcpServers": {
"stillonline": {
"command": "npx",
"args": ["-y", "stillonline-mcp"],
"env": {
"STILLONLINE_API_KEY": "sk_live_YOUR_KEY"
}
}
}
}Client | Config file (examples) |
Cursor |
|
Claude Desktop (Windows) |
|
Claude Desktop (macOS) |
|
Optional env: STILLONLINE_API_BASE (default https://api.stillonline.tech/v1).
Commands
Command | Purpose |
| MCP server (stdio, for editors) |
| Interactive installer |
| CLI help |
Tools
Tool | REST | Notes |
|
| |
|
|
|
|
| Soft-delete |
|
| |
|
|
|
|
| Pause: |
|
| |
|
| No API key |
|
|
|
|
|
Legacy underscore names such as list_projects and create_check are accepted as call aliases, but new clients should use the dot-notation names above.
Rate limits: 120 GET / 30 write per minute per key. On 429, errors include RATE_LIMIT and retry_after.
Development
git clone https://github.com/shenwell/stillonline-mcp.git
cd stillonline-mcp
npm install
npm run build
npx stillonline-mcp setup --dry-runFor local MCP server mode, point your client at node + absolute path to dist/cli.js with env.STILLONLINE_API_KEY.
Maintainers syncing from the private StillOnline monorepo: see PUBLISH.md.
Official MCP Registry
This package is published to the Official MCP Registry as io.github.shenwell/stillonline (npm stdio transport). Listed on aggregators such as MCP.so, Smithery, and PulseMCP after registry ingest.
License
MIT — see LICENSE.
Available Tools
10 toolschecks.createCreate CheckA
Add a URL check to an existing StillOnline project. Use this for additional SaaS endpoints, API health URLs, MCP backends, or autonomous agent health checks.
Inputs: url (required), project_id (required), name, interval_seconds. Returns: check
checks.deleteDelete CheckADestructive
Permanently delete a StillOnline check. Use only when an endpoint should no longer be monitored.
Inputs: check_id (required). Returns: deleted
checks.listList ChecksARead-onlyIdempotent
List HTTP and SSL checks for a StillOnline project. Use this before updating, pausing, deleting, or auditing monitor coverage.
Inputs: project_id (required). Returns: checks
checks.updateUpdate CheckA
Update a StillOnline check. Use this to pause or resume monitoring, rename a check, or change its probe interval after deploy or infrastructure changes.
Inputs: check_id (required), name, enabled, interval_seconds. Returns: check
incidents.createCreate IncidentA
Open an incident on a StillOnline status page. Use this when an outage or degraded service should be visible to subscribers and stakeholders.
Inputs: title (required), status_page_id (required). Returns: incident
incidents.getGet IncidentARead-onlyIdempotent
Get a StillOnline incident by ID. Despite the historical name, this returns one incident record for follow-up status updates or review.
Inputs: incident_id (required). Returns: incident
projects.createCreate ProjectA
Create a StillOnline project with an automatic status page and initial HTTP URL check. Call this after deploying a new SaaS app, API, MCP backend, or AI agent health endpoint that needs uptime monitoring.
Inputs: url (required), name (required), description. Returns: project
projects.deleteDelete ProjectADestructive
Soft-delete a StillOnline project. Removes it from the dashboard and disables checks, matching the UI delete behavior.
Inputs: project_id (required). Returns: deleted, project_id
projects.listList ProjectsARead-onlyIdempotent
List StillOnline projects for the authenticated account. Returns project IDs, names, status page metadata, and plan-scoped project data for follow-up check or incident operations.
Inputs: no named parameters. Returns: projects
status.getGet Public StatusARead-onlyIdempotent
Get public StillOnline status JSON by status page slug. Use this to inspect customer-facing uptime without exposing private account data.
Inputs: slug (required). Returns: Public StillOnline status payload for the requested status page, including page…
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 10 tools
Each tool targets a distinct resource (checks, incidents, projects, status) and action (create, delete, list, update, get), with no overlap. Descriptions clearly differentiate purposes.
All tools follow a uniform object.action naming pattern (e.g., checks.create, incidents.get). No mixing of conventions.
10 tools is well-scoped for a monitoring server, covering the main operations without being overwhelming or too sparse.
CRUD is covered for checks and projects, but incidents only have create and get (missing update/delete). Minor gaps exist but core workflows are supported.
Maintenance
Related MCP Connectors
MCP server for Uptrack uptime monitoring. Manage monitors and incidents from AI agents.
Uptime, SSL, DNS and domain monitoring you can talk to from Claude or any MCP client.
Manage uptime monitors, incidents and status pages from any MCP client.
Health checks for public MCP servers: status, latency, tools, and uptime.
Related MCP Servers
AlicenseNot gradedqualityCmaintenanceMCP server for Uptrack uptime monitoring. Manage monitors and incidents from AI agents like Claude, ChatGPT, and Cursor.13 npmMIT
DevHelm MCP Serverofficial
AlicenseCqualityAmaintenanceMCP server for uptime monitoring, incidents, alerting, and dependency status.12940 PyPI1MIT- AlicenseAqualityAmaintenanceMCP server for Uptime Kuma that enables monitoring and management of uptime monitors, heartbeats, notifications, tags, and maintenance windows via natural language.36702 npm70MIT
- AlicenseAqualityDmaintenanceAn MCP server for Updown.io that enables managing website monitoring checks, viewing downtimes and metrics, configuring alert recipients, and controlling status pages through natural language.156 npmMIT