Onboard your AI coding agent to Uptimely
Paste the sentence below into any coding agent. It fetches /agent-setup/prompt.md and configures itself: it registers the Uptimely MCP server over OAuth in your browser, and only asks for an API key if you also want the SDK or the REST API.
Or set it up by hand
Pick your agent for the exact command or config entry.
Claude Code
One command in a terminal.
Codex
One command in a terminal.
Cursor
Add an entry to ~/.cursor/mcp.json (or .cursor/mcp.json in one project).
OpenCode
Add an entry to opencode.json (project) or ~/.config/opencode/opencode.json.
Windsurf
Add an entry to ~/.codeium/windsurf/mcp_config.json.
VS Code and GitHub Copilot
Add an entry to .vscode/mcp.json.
Gemini CLI
One command in a terminal.
Claude (claude.ai)
Configured in the client's own settings.
ChatGPT
Configured in the client's own settings.
One setup, three ways in
- Remote MCP server
https://app.getuptimely.com/api/mcp - TypeScript SDK
@uptimely/sdk - REST API
https://app.getuptimely.com/v1
The MCP server authenticates with OAuth 2.1 in your browser — no key needed. The SDK and REST API need a project API key from Settings → API keys. Full tool list and write-access rules are in the MCP and AI docs.
Confirm it worked
Ask your agent to call search_tools, then execute_typescript with return await external_uptimely_project_list({});. A client that lists uptimely_project_list as its own tool can call it directly. A JSON list of your projects — even an empty list — means the connection is live. Each guide below also carries a REST check for the SDK/API path.
Canonical copy of the machine-readable instructions: https://getuptimely.com/agent-setup/prompt.md