Use with AI Agents

Qirabot isn't only a library you code against — it's a capability an AI agent can pick up and use. Tell your agent "automate this signup flow and verify the confirmation email screen" and let it write and run the script. Two ways in:

Claude Code plugin

The qirabot plugin packages an Agent Skill that teaches Claude Code how to operate Qirabot end to end. Install it once:

/plugin marketplace add qirabot/claude-plugins
/plugin install qirabot@qirabot

Claude then invokes the skill automatically whenever a task involves automating, testing, or scraping a UI (it's also callable explicitly as /qirabot:qirabot). The skill ships with:

• A preflight script — checks the Python env, backend deps, and API key before writing any code, and prints the exact fix command for anything missing. No more scripts that fail three steps in.
• Condensed SDK + CLI references — the agent codes against an accurate API surface instead of guessing.
• Starter templates for browser, Android (adb), iOS (WDA and Appium), and bring-your-own-driver bolt-on — the agent adapts a working skeleton rather than starting blank.

The plugin contains instructions and helpers only; the qirabot package itself is installed at runtime by the preflight bootstrap.

Any other agent (Cursor, Copilot, …)

Two properties make Qirabot easy for arbitrary agents to drive:

The CLI is a natural agent tool. One shell command runs a whole natural-language task — no code generation needed for one-shot jobs:

qirabot browser "Fill the signup form as Jane Doe and stop at the captcha" --url example.com

Exit codes are machine-checkable (0 pass, 1 fail, 130 interrupted), and every run writes an HTML report with per-step screenshots the agent (or you) can inspect when something goes wrong. All commands: CLI Reference.

The docs are agent-readable. Point your agent at:

• https://qirabot.com/docs/llms.txt — index with per-page summaries
• https://qirabot.com/docs/llms-full.txt — the complete docs in one file
• every page as raw Markdown by swapping .html for .md, e.g. https://qirabot.com/docs/reference/methods.md

In Cursor, add them via the @Docs feature; in other tools, reference the URLs in your rules file or paste them into context.

Why agents + vision automation fit

An agent writing Playwright still has to guess selectors it can't see, and they break on the next redesign. With Qirabot the agent describes elements the way it reasons — in language — and the same skill covers surfaces code-first stacks can't reach: native mobile apps, desktop software, and games. See What is Qirabot and the platform support matrix.