AI onboarding

Onboard your AI agent to Keystroke.

Keystroke is designed for building with your coding agent. This page helps your coding agent get set up with Keystroke. For a step-by-step walkthrough you run yourself, see the Quickstart.

Recommended: a local project plus the CLI for agents. That is the default we document and the most stable loop for agents in an editor. The Keystroke MCP server is an alternative for chat-first agents with no local install — see Build over MCP instead below.

You can also build without a coding agent entirely: the platform agent is a built-in building agent in the Keystroke web app. This page is about onboarding your own coding agent; the platform agent needs no onboarding.

Give your agent the setup prompt

The fastest path to setting up Keystroke is giving your agent the following prompt:

Read https://keystroke.ai/start.md then help me create my first agent.

What your agent gets

keystroke init scaffolds a project that’s ready for a coding agent from the first commit:

File	What it gives your agent
AGENTS.md	Baseline context: the CLI, project layout, discovery, and working habits
CLAUDE.md	Symlink to AGENTS.md so Claude Code picks it up automatically
tsconfig.json	Committed TypeScript config (extends @keystrokehq/cli/tsconfig.json)
package.json	Scripts for build, lint, typecheck, and test
src/	An example agent to read and extend

Beyond this baseline, your agent reads the full docs on demand with keystroke docs search and keystroke docs query — see docs for agents.

Most coding agents read AGENTS.md automatically when they open the project, so they’re oriented before you ask for anything.

Start building

With the project open in your coding agent, just describe what you want. For example:

Build a morning brief agent. It should scan my Google Calendar, Gmail inbox, and Linear issues each weekday morning and send me a short briefing in Slack.

Build a meeting recap agent. It should trigger whenever one of my Google Meet calls finishes, turn the transcript into action items, and send me a recap in Slack.

Build over MCP instead

Use this path only when local CLI setup is not an option — for example, a chat-first agent in ChatGPT or Claude with no repo on disk.

Connect the Keystroke MCP server. Your agent works in a hosted workspace with the same docs access and deploy loop, so you skip keystroke init entirely. The MCP hosted-workspace experience is newer than the CLI; when your agent can run in an editor locally, start with the setup prompt above instead.

Give your agent the docs

This page is often the first thing an agent reads. You can hand it Keystroke’s docs in a context-aware way:

1. All the docs (llms.txt)

   Point your agent at the complete, machine-readable docs index:

   Here are the Keystroke docs: https://keystroke.ai/docs/llms.txt
   

2. A single page as markdown

   Append .md to any docs URL to get a clean markdown version:

   Docs for this page: https://keystroke.ai/docs/build-with-ai/onboarding.md
   

3. In your editor

   Every docs page can be copied or opened directly in ChatGPT, Claude, Cursor, or an MCP client from the menu at the top of the page.