Set up with AI

Qikbar is designed so that a coding agent can integrate it in one pass. The agent reads your router, pages and API, then writes the config. Copy this prompt into Claude Code, Cursor, Windsurf, GitHub Copilot, Codex, or any other agent:

You can also fetch it directly: /prompt.txt.

Documentation for AI

Every page in these docs is available as plain Markdown:

• /llms.txt: an index of every page, following the llms.txt convention
• /llms-full.txt: the entire documentation in one file
• Any page + .md: for example /docs/sources.md

Each docs page also has Copy as Markdown, Open in Claude and Open in ChatGPT buttons at the top.

Tips for agents

• Mount Qikbar once. In React, render <Qikbar> near the root. Elsewhere, call createQikbar in your client entry.
• Use the app’s router for navigate. Never hard-code window.location if a router exists.
• Use sources for data that lives on the server, with the app’s existing fetch layer.
• Theme with the app’s existing CSS variables, for example vars: { accent: 'var(--primary)' }.
• Keep AI keys on the server with createHandler from @qikbar/qikbar/server.

Add Qikbar to your project instructions

If your repo has an AGENTS.md, CLAUDE.md or .cursor/rules, add this line so future changes keep the bar up to date:

When adding a page, route, or user-facing action, also register it with Qikbar (see src/command-bar.ts). Qikbar docs: https://qikbar.com/llms.txt