AI Agents & Coding Assistants

From the beginning, Checkly has bet on Monitoring as Code which lets you create and control your monitoring infrastructure entirely using code. By default, Checkly constructs reflect all your monitoring properties.

api.check.ts

import { ApiCheck, AssertionBuilder } from "checkly/constructs"

new ApiCheck("api-health-check", {
  name: "API Health Check",
  request: {
    url: "https://danube-web.shop/api/books",
    method: "GET",
    assertions: [
      AssertionBuilder.statusCode().equals(200),
    ],
  },
})

All your monitoring resources can be updated, tested and deployed via the Checkly CLI.

# test your monitoring configuration
npx checkly test

# deploy and update your monitoring setup
npx checkly deploy

The Monitoring as Code workflow is by default AI-native because LLMs are excellent at writing and editing Checkly constructs code and modern AI agents can execute CLI commands easily. Provide the necessary Checkly context and let your AI agent of choice do the rest.

Create new checks, alert channels or other constructs

“Can you set up a new BrowserCheck for example.com?”

Bulk-update your monitoring resources

“Can you change all checks to run every 5 minutes instead of every 10 minutes?”

Gather information about alerts and your monitoring setup

“I just received an alert. Can you tell me details about the failing checks?”

Handle and communicate incidents

“Can you please open an incident and investigate a fix?”

Add Checkly context to your AI agent conversation

Install Checkly Skills or connect the Checkly MCP Server to give your AI agent enough context to perform Checkly-related tasks.

Checkly Skills

For coding agents with repo and code access. Author Monitoring as Code with the Checkly CLI.

Checkly MCP Server

For desktop or sandboxed agents. Live access to Checkly account data and remote actions.

Skills, MCP, and the CLI

Use Skills for a CLI-first workflow. Skills are built for coding agents that have access to your repository and can run commands. The agent edits your Checkly constructs and tests, then uses the Checkly CLI to test and deploy them. Skills load context on demand, keeping your agent’s context window lean until Checkly-related tasks arise. This is the recommended approach for agents that support the Agent Skills standard. Use the MCP Server in desktop or sandboxed environments. Some agents, like Claude Desktop, ChatGPT, or a sandboxed assistant, can’t access your repository or filesystem. The Checkly MCP Server connects these clients to Checkly over Streamable HTTP and gives them live data and remote account actions, such as reading check status, inspecting test sessions, triggering existing checks, reading RCA, or managing status page incidents.

Use case	Recommended path
Ask quick, ad hoc questions from a supported client	MCP Server
Inspect live Checkly data without local CLI setup	MCP Server
Trigger existing deployed checks from chat	MCP Server or Checkly CLI
Create or edit check code	Checkly Skills with the Checkly CLI
Test, deploy, or automate local Monitoring as Code projects	Checkly CLI
Add reusable Checkly best practices to an agent	Checkly Skills

Feed documentation to your agent

Every Checkly documentation page is available as markdown, and an llms.txt index lists them all. Both make it easy to pull the exact docs your agent needs into its context.

Markdown Access

Append .md to any docs URL, request markdown with content negotiation, or copy a page as markdown.

llms.txt

A machine-readable index of every documentation page for crawling and indexing.

Let Checkly’s AI run the monitor

The tools above help your AI agent author Monitoring as Code. If you’d rather describe a monitoring goal in plain language and let Checkly’s AI run it for you, use Agentic Checks. They turn a prompt into a synthetic check that explores your app, evaluates assertions, and self-heals when the underlying flow changes.

Agentic Checks

Turn a prompt into an AI-powered synthetic check that discovers, verifies, and maintains itself.