AdsCrawl provides a browser for AI agents through its hosted MCP server, Remote CDP sessions and browser task endpoints. Use MCP with a compatible agent client, or connect Playwright to build your own tools. Planning, model inference and approval rules remain in the agent application.
Remote CDP Browser: Use temporary Chromium sessions as the execution layer for the agent's interactive browser tools.
Cloud Browser API: Evaluate persistent browser profiles when work must resume with saved state. This is a separate lifecycle from the agent's CDP example.
Which browser setup should an AI agent use for authenticated tasks?
Use a temporary Remote CDP session for actions completed within one task, and evaluate a saved Cloud Browser profile when work must resume with retained cookies and settings. These are separate AdsCrawl products: the saved profile is not automatically available to the CDP example below. Neither profile reuse nor an anti-detection feature guarantees access, permanent login or successful agent execution.
Test the required login and MFA flow with an authorized account, then verify the state after the exact stop, restart or reconnect steps you need.
For persistent Cloud Browser work, allow for a paid runtime, an explicit proxy on API-key starts and login-session access to the interactive viewer.
Keep planning in your agent application; measure valid task completion and cleanup on your own sites before comparing reliability or deployment speed.
AdsCrawl MCP lets an agent client that supports Streamable HTTP open a browser, read rendered pages and interact with them. It uses temporary Remote CDP sessions; your agent still handles task planning.
Browser execution endpoint
https://api.adscrawl.net/mcp
Transport
Streamable HTTP
Authentication
x-api-key — store in your client's secret settings
Create an API key in the AdsCrawl dashboard and add the endpoint as a remote MCP server.
Connect, list tools, then ask your agent to open a public test page, read its heading and close the session.
browser_open → browser_read → browser_close. Use browser_navigate if needed; check for confirmed: true in the close result.
Open https://example.com in an AdsCrawl browser on the US route, read the page heading, then close the browser and confirm it stopped.
The expected heading is Example Domain. browser_open starts a billed session; close it after completion or an error. browser_screenshot publishes a JPEG that anyone with its link can view, so obtain permission before publishing.
Free includes 500 one-time credits for supported temporary sessions; starting a persistent Cloud Browser profile requires a paid plan. MCP sessions share your API key's concurrency limit. For HTTP 429, release a session or wait for capacity.
https://docs.adscrawl.net/mcp is a separate documentation search service. Use the api.adscrawl.net endpoint above for browser interaction.
A recorded MCP pagination run
On 2026-11-12, an MCP run opened the JavaScript Quotes to Scrape test page on the US route, read the page, clicked Next and checked its content:
Initial page — 10 author entries; Albert Einstein.
Read immediately after pagination — 0 author entries.
Read again after waiting — 10 author entries; Marilyn Monroe.
The first page-two read arrived before its content rendered; a follow-up read found 10 author entries. Cleanup returned confirmed: true with status stopped. Check that the content you need is ready, not just that the URL changed.
One public-page sample does not establish speed, success rates or persistent-profile behavior. The download retains timestamps, URLs, counts, first authors and cleanup status, with session identifiers, tokens and quote bodies omitted. Download the recorded trace ↓
A recorded Playwright run against an AdsCrawl Remote CDP session: open the JavaScript page, follow its Next link, read page two, then delete the session.
The URL changes to /js/page/2/ and the first author changes from Albert Einstein to Marilyn Monroe. The run ends with a successful session DELETE response.
A recorded public-page example. Results reflect the page at capture time and can change with the target website, loading conditions and request settings.
02 / USE CASES
Where it fits
01
Observe the current page
Read page text or capture a screenshot using your browser client, then pass the required observation to your model. Keep the page URL with the result.
02
Execute bounded actions
Map model-selected actions to explicit browser tools such as navigation or a known field interaction. Validate inputs and inspect the page after each action.
03
Choose a lighter tool when possible
Use Markdown or structured extraction for read-only content tasks. Use a CDP session when a task needs multiple interactive steps.
03 / QUICKSTART
Start with one complete workflow
Choose your agent framework and keep the AdsCrawl API key in its server-side tool layer.
Create a Remote CDP session and connect Playwright; the example below demonstrates the browser connection, not an autonomous agent.
Implement your observe, plan and execute loop with validated tools, step limits and approval rules for consequential actions.
Save the needed results and delete the remote session in cleanup, including after model or tool failures.
Requires Node.js and Playwright (npm install playwright). Set the ADSCRAWL_API_KEY environment variable on your server first.
AdsCrawl supplies browser execution, not a hosted LLM, agent planner or guaranteed task completion.
Web content is untrusted input. Keep page instructions separate from your agent's permissions and user intent.
Model tokens and browser usage have separate costs. Set action budgets and session timeouts in your application.
Framework integrations must support the chosen CDP or HTTP tools. This page does not imply a native integration with every agent framework.
Keep API keys on the server. For HTTP 402 check your balance; for 429 reduce concurrency and retry with backoff.
05 / FAQ
Common questions
Does the API accept a goal and finish the task itself?
The APIs shown here perform browser operations. Your application must implement planning, tool execution and completion checks.
Can I connect an AI agent through MCP?
Yes. Use https://api.adscrawl.net/mcp with Streamable HTTP and an x-api-key header stored in your client's secret settings. Discover the browser tools, open and read a page, then call browser_close and check for confirmed: true. The separate docs.adscrawl.net MCP searches documentation.
Do I always need a running browser session?
No. For readable content, use the Markdown output of POST /html. Choose Remote CDP for interactive multi-step workflows.
Can the agent reuse a session?
It can perform multiple operations within an active session's lifetime. Delete the session after the task; use saved profiles separately when persistence is needed.
ADSCRAWL PRODUCTS
Related workflows
Choose the next workflow by the output, client or browser state your task needs.