Agent-Native Synthetic Monitoring & Status Pages

Synthetic Uptime & SLAs Built for AI Coding Agents

Equip Claude, Gemini, Cursor, and developer teams with automated synthetic probes, dead-man switch heartbeats, and white-labeled status pages. 100% Cloud-Native Serverless.

Run Instant Synthetic Check from Global Edges

Test Any Website or API Endpoint Live

Multi-Region Edge Ready
Core Capabilities

Synthetic Checks, Cron Heartbeats, and Status Pages

Model Context Protocol (MCP)

Seamlessly interact with Claude, Gemini, Cursor, or AutoGPT via Stdio and SSE. Provision checks with simple agent tool calls.

Dead-Man Heartbeats

Monitor background agent tasks, cron jobs, and batch queues with unique ping tokens and configurable grace periods.

Branded Status Pages

Publish white-labeled status pages with custom domains, 90-day history bars, incident timelines, and RSS feeds.

Agent Onboarding — 2 minutes, no dashboard spelunking

Let your coding agent set up Seseli for you

Copy one prompt, click one deep link, give consent. Your agent handles the rest — monitors, live checks, and a status page on status.<your-domain>.

1 · Copy the prompt

Copy the prompt below and paste it into your coding agent (Claude Code, Cursor, Gemini, etc.). It tells the agent to connect to the Seseli MCP server and follow the onboarding sequence.

2 · Open the deep link + consent

Your agent will call start_onboarding and show you a deep link. Open it, sign in or create a free account, then click "Authorize agent". That issues a scoped ssl_live_... key — paste it back to your agent (never share your password).

3 · Agent sets up your monitoring

Once authenticated, your agent creates the monitor, runs a live check, asks if you want status.<your-domain> (full DNS walkthrough included), and publishes your public status page.

1 · Copy this prompt to your agent

The agent connects to Seseli MCP, shows you the consent deep link, and walks through monitoring + status.<domain> DNS setup.

Set up website monitoring and a public status page for my site with Seseli via MCP.

HOW TO CONNECT TO SESELI MCP (use whichever your client supports):
- Local / stdio: run its MCP server with: npx -y @seseli/mcp-server
- Remote / Streamable HTTP: POST https://api.seseli.io/v1/mcp (JSON-RPC 2.0).
  Discover first with GET https://api.seseli.io/.well-known/mcp.json —
  then call tools/list.

  No API key yet? No problem — get the deep link directly (no auth needed):
  POST https://api.seseli.io/v1/onboarding/deep-link
  with {"siteUrl": "https://example.com", "agent": "Claude Code"}
  (stdio MCP + start_onboarding also work without a key).

SEQUENCE — follow this exactly:
1. Call the Seseli MCP tool `start_onboarding` (or read resource
   `seseli://onboarding/guide` / prompt `onboarding` if you prefer), passing
   my website URL in `siteUrl` if I already gave it to you. The tool returns
   a deep link (https://app.seseli.io/onboarding?... on prod, or the dev/stage
   host in those environments). If remote MCP returns 401/Unauthorized because
   I have no key yet, fall back to POST /v1/onboarding/deep-link above, or
   just give me https://app.seseli.io/onboarding directly.
2. Show me that deep link as a clickable link and explain: open it on your
   phone/computer, sign in (or create a free account if you don't have one),
   then click "Authorize agent". That consent step issues a scoped
   ssl_live_... agent API key. STOP and wait — do not continue until I confirm
   I have consented and paste the key (or set SESELI_API_KEY) for you.
   Never ask for my password; the key is the only credential I should share.
3. Once authenticated, set up monitoring for my website:
   a. If you don't know my website URL yet, ask me for it.
   b. Call `create_monitor` with a sensible name, the URL, intervalSeconds 300,
      and tags ["agent-onboarded"].
   c. Call `trigger_check_now` to verify it is reachable and report
      status, latency, and TLS status in plain language.
4. ASK ME: "Do you want a public status page at status.<my-domain>?"
   - If yes, call `create_status_page` with a slug from my domain,
     title "<My Domain> Status", all monitor IDs you just created, and
     customDomain "status.<my-domain>". Then walk me through DNS step by step:
     add a CNAME record — Host "status" → Value "status.seseli.io" — at my
     registrar (give me Cloudflare/Route 53/Namecheap/GoDaddy wording if you
     can tell which one I use), how to verify with `dig CNAME`, propagation
     timing, and note the TLS certificate is provisioned automatically.
     Show me the public URL https://status.seseli.io/<slug> and the target
     https://status.<my-domain> (after DNS propagates).
   - If no, create the status page without a custom domain and show me the
     public https://status.seseli.io/<slug> URL.
5. Summarize what you did: monitor name + URL + check result, status page
   public URL (+ custom domain + DNS status if requested), and how to view it
   in the dashboard at https://app.seseli.io/dashboard.

RULES:
- Keep me in the loop: tell me before each step, pause at the consent step,
  and never create monitors or pages without my confirmation.
- If no MCP server is connected, tell me the exact 2-minute fix (add
  "seseli" with command "npx", args ["-y", "@seseli/mcp-server"] to my MCP
  client config) and wait for me to reconnect.
- If a step fails, explain the error in plain language and offer the next
  recovery step. Do not print secret keys back in full once set.
Connect your agent (stdio — recommended for local agents):
npx -y @seseli/mcp-server

Or remote (cloud agents, no install): https://api.seseli.io/v1/mcp · discovery https://api.seseli.io/.well-known/mcp.json

What your agent will show you (deep link):
https://app.seseli.io/onboarding?agent=Your+Coding+Agent
Preview the consent page

Opening the deep link signs you in (or creates an account), then asks for consent to issue a scoped agent key. Your agent only ever sees the key — never your password.

Want status.<your-domain>? During onboarding your agent will ask, create the status page with that custom domain, and walk you through the single CNAME record (status → status.seseli.io) for your registrar — including how to verify it with dig CNAME.

1-Click Model Context Protocol (MCP) Setup Claude Desktop & Gemini Ready

Add Seseli to your AI Agent or Cursor / Claude Desktop environment in seconds.

claude_desktop_config.json / mcp.json
{
  "mcpServers": {
    "seseli": {
      "command": "npx",
      "args": [
        "-y",
        "@seseli/mcp-server"
      ],
      "env": {
        "SESELI_API_KEY": "ssl_live_YOUR_API_TOKEN_HERE"
      }
    }
  }
}
Or run directly via CLI / Agent Shell:
npx -y @seseli/mcp-server
Supported MCP Workflows
  • Deploy and verify synthetic monitors right after generating backend code.
  • Inspect latency SLA breaches and automatically post incident status updates.
  • Set up dead-man switches for autonomous background jobs and scheduled tasks.
  • Query live 90-day uptime and SSL certificates for compliance audits.

Interactive MCP Tool Execution Simulator

Simulate how Claude, Gemini, or autonomous agents execute Seseli MCP JSON-RPC tool calls.

Tool Input (JSON Arguments):
Agent Tool Output (JSON-RPC Result):
// Tool response will render here...
Infrastructure

Usage-Based Serverless Execution

Multi-region HTTP probes with DNS/TCP/TLS waterfall timings, KMS-encrypted telemetry storage, and HMAC-signed audit logs.

Idle cost
$0.00 / mo
Scale to Zero
Probe latency
< 15ms
Edge Acceleration
Global edges
450+ Points
Global Edge CDN
Compliance
SOC 2 & HIPAA
Hardware KMS + Audit Logs