> ## Documentation Index
> Fetch the complete documentation index at: https://intunedhq.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Intuned Agent overview

## What is Intuned Agent?

Intuned Agent is an AI agent that builds, edits, and maintains browser automation Projects. It runs inside the Intuned platform with a real browser, full access to the platform via the [Intuned CLI](/main/05-references/cli/overview), and works in the background as long as needed.

## Where to find the agent

Access the agent at [app.intuned.io/agent](https://app.intuned.io/agent). Chat with the agent, review its plan, approve changes, and deploy — all from one session.

<Frame caption="Sped-up demo of Intuned Agent creating a scraper.">
  <video autoPlay muted loop playsInline controls src="https://mintcdn.com/intuned-dev/PKoj-Mu2mK0gzdqc/assets/intuned-agent/intuned-agent-overview.mp4?fit=max&auto=format&n=PKoj-Mu2mK0gzdqc&q=85&s=7b74d936064516e240155b714d0047ea" data-path="assets/intuned-agent/intuned-agent-overview.mp4" />
</Frame>

## What the agent can do

<AccordionGroup>
  <Accordion title="Create scrapers, RPAs, crawlers, and AI-powered projects" icon="hammer">
    * Scrapers for any website — public, login-protected, bot-protected,
      CAPTCHA-gated - RPA automations — form submissions, data entry, account
      actions - Crawlers — multi-page discovery, sitemaps, content indexing -
      AI-powered automations — [Stagehand](/main/04-integrations/stagehand) /
      [Browser-Use](/main/04-integrations/browser-use)
  </Accordion>

  <Accordion title="Handle AuthSessions, CAPTCHAs, bot detection, and complex pages" icon="shield-halved">
    * Pagination, infinite scroll, load more, detail pages, nested lists,
      iframes - [AuthSessions](/main/02-features/auth-sessions) — secure
      credential collection, auth session setup - [Stealth, CAPTCHA, and
      proxies](/main/02-features/stealth-mode-captcha-solving-proxies) - Network
      interception — prefer XHR/GraphQL over DOM scraping
  </Accordion>

  <Accordion title="Fix broken automations — selectors, timeouts, and failed runs" icon="wrench">
    * Fix from a failed [Run](/main/02-features/runs-single-executions) — broken
      selectors, changed page structure, timeouts, wrong output - Fix from a
      failed [JobRun](/main/02-features/jobs-batched-executions) — investigate
      issues across batched executions - Debug with Playwright trace analysis
  </Accordion>

  <Accordion title="Detect, fix, and redeploy automatically with self-healing projects" icon="rotate">
    Enable self-healing on any project and the agent monitors runs, detects
    issues, writes fixes, and redeploys — without manual intervention. See
    [Self-healing projects](/main/02-intuned-agent/self-healing-projects).
  </Accordion>

  <Accordion title="Deep platform integration" icon="plug">
    The agent isn't an external wrapper — it reads Run and JobRun results,
    traces, and logs, configures
    [Jobs](/main/02-features/jobs-batched-executions), and tests on the same
    infrastructure your Project uses in production.
  </Accordion>
</AccordionGroup>

## Agent in the online IDE

The same agent is also available in the [Intuned online IDE](/main/02-features/online-ide) Chat tab. Use it to make quick edits to an existing Project directly in the code editor, without opening a full agent session.

<Frame>
  <video autoPlay muted loop playsInline controls src="https://mintcdn.com/intuned-dev/PKoj-Mu2mK0gzdqc/assets/intuned-agent/ai-in-ide.mp4?fit=max&auto=format&n=PKoj-Mu2mK0gzdqc&q=85&s=12a99b8393ae768e2f736803cf7f1ec7" data-path="assets/intuned-agent/ai-in-ide.mp4" />
</Frame>

## Cost and billing

The agent consumes AI credits based on tokens used. Free plans include credits to get started. Track spend on the [Usage page](https://app.intuned.io/usage) under the **IntunedAgent** project.

See [Plans and billing](/main/05-references/plans-and-billing) for plan limits and [Usage and billing](/main/02-features/usage-pricing-monitoring-alerting) for tracking spend. You can also [bring your own Anthropic API key (BYOK)](/main/02-features/usage-pricing-monitoring-alerting#bring-your-own-key-byok) instead of using AI credits.

## FAQ

<AccordionGroup>
  <Accordion title="Can I review the code before deploying?">
    Yes. Review and iterate on the plan with **Suggest Changes**, or select **Approve Plan**. After the agent builds, inspect the generated code in the **Code** panel and test results in the **Artifacts** panel. When ready, select **Merge Branch** — choose **Merge & Deploy** to deploy immediately, or **Merge only** to merge without deploying.
  </Accordion>

  {" "}

  <Accordion title="Can the agent bypass CAPTCHAs or bot detection?">
    Yes. The agent enables [stealth mode and CAPTCHA
    solving](/main/02-features/stealth-mode-captcha-solving-proxies) based on what
    it encounters — stealth mode for bot detection, CAPTCHA solving if a challenge
    is still visible after stealth, and proxies for IP-based blocking.
  </Accordion>

  {" "}

  <Accordion title="Can the agent work with authenticated websites?">
    Yes. The agent securely collects credentials (never visible to the AI) and
    sets up [AuthSession](/main/02-features/auth-sessions) login flow APIs.
  </Accordion>

  {" "}

  <Accordion title="Can the agent suggest a schema?">
    Yes. The agent auto-discovers the schema by exploring the site and analyzing
    the data available on the page. It includes the proposed schema in the plan
    for your review.
  </Accordion>

  {" "}

  <Accordion title="Why does the agent take so long?">
    The agent explores and understands the site, generates and tests code, then
    runs a full validation. Complex automations with pagination, edge cases, and
    large data sets take longer.
  </Accordion>

  {" "}

  <Accordion title="Can I close my browser while the agent is working?">
    Yes. The agent continues working in the background. Close your browser or
    navigate away — come back anytime to check progress or review results.
  </Accordion>

  <Accordion title="How is this different from Director.ai?">
    Both tools use AI to generate browser automation code, but they differ in execution:

    |                     | Intuned Agent                                          | Director                                 |
    | ------------------- | ------------------------------------------------------ | ---------------------------------------- |
    | **Execution**       | Deterministic code by default, AI-driven code optional | AI agent controls the browser at runtime |
    | **Data extraction** | Code-based selectors — predictable, fast               | LLM-powered — flexible, slower           |
    | **Output**          | Complete Playwright project                            | Stagehand scripts                        |
    | **Platform**        | Deploys to Intuned with jobs, scheduling, and auth     | -                                        |

    Intuned Agent produces deterministic code by default. It can also generate AI-driven automations using [Stagehand](/main/04-integrations/stagehand) or [Browser-Use](/main/04-integrations/browser-use) when requested.
  </Accordion>

  {" "}

  <Accordion title="Does the agent support CLAUDE.md and custom skills?">
    Not yet, but coming soon. The agent will load your project's CLAUDE.md and
    custom skills to understand your project's structure, conventions, and
    extended capabilities.
  </Accordion>

  {" "}

  <Accordion title="Does the agent work on a branch?">
    Yes. Each session creates an auto-generated branch prefixed with `agent/`
    (e.g. `agent/fix-amazon-scraper-timeout`). Your changes are isolated until you
    merge. View branches from the project's overview page — both new and existing
    projects.
  </Accordion>

  {" "}

  <Accordion title="Can I run multiple sessions at once?">
    Yes. You can run concurrent agent sessions, subject to your workspace's
    concurrency quota. See [Plans and
    billing](/main/05-references/plans-and-billing) for limits.
  </Accordion>

  {" "}

  <Accordion title="What languages does the agent support?">
    TypeScript and Python. The language is chosen when creating a new project.
    When editing an existing project, the agent matches the language of the
    existing code.
  </Accordion>

  {" "}

  <Accordion title="Can the agent install dependencies?">
    Yes. The agent manages `package.json` (TypeScript) or `pyproject.toml`
    (Python) and can add or remove packages as needed.
  </Accordion>

  {" "}

  <Accordion title="Can the agent manage environment variables?">
    Yes. The agent uses a secure tool to request environment variables and secrets
    — it never asks you to paste sensitive values into the chat.
  </Accordion>

  {" "}

  <Accordion title="Can I use my existing Claude Code subscription?">
    No. Intuned Agent uses its own AI credits system, separate from any Claude
    Code or Anthropic Console subscription. You can [bring your own Anthropic API
    key
    (BYOK)](/main/02-features/usage-pricing-monitoring-alerting#bring-your-own-key-byok)
    to use your own key instead of AI credits.
  </Accordion>

  <Accordion title="Can the agent run locally?">
    Not yet. A local plugin for coding agents is coming soon, allowing you to run the agent from your own development environment.
  </Accordion>
</AccordionGroup>

## Guides

<CardGroup cols={2}>
  <Card title="Using the agent" icon="book" href="/main/02-intuned-agent/using-the-agent">
    Workflows, prompt examples, and capability deep dives
  </Card>

  <Card title="Self-healing projects" icon="rotate" href="/main/02-intuned-agent/self-healing-projects">
    Automated fix, merge, and deploy pipelines
  </Card>

  <Card title="Quickstart: Scraper" icon="database" href="/main/00-getting-started/quickstarts/intuned-agent">
    Build a scraper with the agent step by step
  </Card>

  <Card title="Quickstart: RPA" icon="robot" href="/main/00-getting-started/quickstarts/build-an-rpa-with-intuned-agent">
    Build an RPA automation with the agent
  </Card>
</CardGroup>