Bright Data connector for AI agents

Q: Does the agent access Bright Data as the user or through a shared token?

As the user. Each user connects their own Bright Data access token once, and Scalekit sends it with every call. Audit logs attribute every scrape job and search to that user, not a shared service account.

Q: Where is the Bright Data access token stored?

In Scalekit's managed AES-256 token vault, namespaced per tenant. To rotate it, the user reconnects with a new access token. Revocation is a single dashboard action. Credentials never appear in prompts, logs, or LLM context.

Q: Can I limit what the agent does in Bright Data?

Yes. Bright Data has no delete tools. Filter by tool name in listScopedTools to leave out tools that start jobs, such as brightdata_trigger_scrape, and brightdata_filter_marketplace_dataset, which creates a billable snapshot. Scalekit also enforces scope checks before every API call.

Q: What happens when a user revokes Bright Data access?

The connection is invalidated on the next tool call. Subsequent requests for that user fail closed with a clear error. Other users in the tenant remain unaffected. The event is logged for audit.

Q: Can the agent check Bright Data spend before a large job?

Yes. brightdata_get_balance returns the account balance, credit, prepayment, and pending costs in USD, and brightdata_get_zone_cost returns billed cost and bandwidth for one zone over a date range.

Tools your web data agent reaches for on Bright Data, scoped per user.

CALL ANY TOOL

43 tools for web scraping with Bright Data: fetch public pages through Web Unlocker, run SERP searches, trigger and download scraper jobs, run Scraper Studio collections, search company and people records, and check zones, usage, and balance.

brightdata_cancel_snapshot

Cancel snapshot

Cancel a running Bright Data scraper job (snapshot). Returns OK when the job is canceled. Use this to stop a job started with brightdata_trigger_scrape. Use brightdata_rerun_snapshot to run a finished job again.

Parameters

Name

Type

Required

Description

snapshot_id

string

Required

Snapshot ID returned by brightdata_trigger_scrape, brightdata_scrape_sync (when it times out into async mode) or brightdata_list_snapshots. Starts with s_ or sd_. Example: s_m4x7enmven8djfqak

brightdata_download_marketplace_snapshot

Download marketplace snapshot

brightdata_download_snapshot

Download snapshot

brightdata_get_account_bandwidth

Get account bandwidth

brightdata_get_account_status

Get account status

brightdata_get_browser_session

Get browser session

brightdata_get_dataset_metadata

Get dataset metadata

brightdata_get_serp_result

Get serp result

brightdata_get_snapshot_input

Get snapshot input

brightdata_get_snapshot_progress

Get snapshot progress

brightdata_get_studio_batch_result

Get studio batch result

brightdata_get_studio_realtime_result

Get studio realtime result

brightdata_get_unlock_result

Get unlock result

brightdata_get_zone

Get zone

brightdata_get_zone_bandwidth

Get zone bandwidth

brightdata_list_active_zones

List active zones

brightdata_list_all_zones

List all zones

brightdata_list_browser_sessions

List browser sessions

brightdata_list_scrapers

List scrapers

brightdata_list_snapshots

List snapshots

brightdata_list_studio_jobs

List studio jobs

brightdata_rerun_snapshot

Rerun snapshot

brightdata_run_studio_realtime

Run studio realtime

brightdata_scrape_sync

Sync scrape

brightdata_search_companies

Search companies

Search Bright Data's Business Search index of company records with a plain-English query or structured field conditions. Returns req_id, meta (matched count, coverage, offset, limit) and a documents array whose data fields depend on view or fields. Use this to find companies without knowing their URLs; use brightdata_search_people for people, and brightdata_scrape_sync to collect a known profile URL.

Parameters

Name

Type

Required

Description

mode

string

Required

How the query is interpreted. instant: fast natural-language search, up to 100 results. smart: natural-language search reranked by relevance, up to 10 results. ludicrous: structured field conditions passed in structured_query, up to 100 results per page. One of: `instant`, `smart`, `ludicrous`.

fields

array

Optional

Exact list of field names to return in each result, instead of a preset view. Unknown field names are rejected with HTTP 400. Example: ["name", "website", "industry", "employees_in_linkedin", "url"]

limit

integer

Optional

Maximum number of results in this page. Server default 10; maximum 100 for ludicrous and instant, 10 for smart. Example: 25

offset

integer

Optional

Number of results to skip, for paging. ludicrous accepts up to 1,000; in instant and smart, offset plus limit cannot exceed 100. Example: 10

query

string

Optional

Plain-English description of the company to find, up to 200 characters. Required when mode is instant or smart; ignored for ludicrous. Example: US software companies with 50 to 500 employees

source

string

Optional

Dataset to search within the company category. Omit to use the default source, linkedin_company. Example: linkedin_company

structured_query

object

Optional

Structured search conditions, required when mode is ludicrous. Combine conditions with and, or and not; each condition uses one field operator: equals for exact matches, in for a list of values, range for numeric bounds (>=, <=) and text for keyword matches. See https://docs.brightdata.com/products/business-search/query-syntax for field names. Example: {"and": [{"text": {"industry": "software"}}, {"equals": {"headquarters_country_code": "US"}}]}

view

string

Optional

Which stored fields each result's data object carries: id_only (the default, identifiers only), summary or full. Ignored when fields is set. One of: `id_only`, `summary`, `full`.

brightdata_search_marketplace_dataset

Search marketplace dataset

brightdata_serp_search

Search serp

brightdata_trigger_scrape

Trigger scrape

Start an asynchronous Bright Data pre-built scraper job (a snapshot) for many inputs, optionally with a discovery phase. Returns snapshot_id for the new job. Use this for large batches or discovery (by keyword, category URL, location and so on). Use brightdata_scrape_sync for a few inputs you need right away. Get the dataset_id and input fields from brightdata_list_scrapers. Then poll brightdata_get_snapshot_progress until status is ready and fetch records with brightdata_download_snapshot.

Parameters

Name

Type

Required

Description

dataset_id

string

Required

Scraper dataset ID that selects the pre-built scraper to run. Look it up with brightdata_list_scrapers or at https://docs.brightdata.com/scrapers.json; never guess it. Example: gd_l1viktl72bvl7bjuj0 (LinkedIn people profiles).

input

array

Required

Array of input objects for the scraper. Each object's fields depend on the scraper and collection method: most collect-by-URL scrapers take a url, while discovery methods take fields such as keyword, location or user_name. Check the scraper's input schema with brightdata_list_scrapers. Example: [{"url": "https://www.linkedin.com/in/satyanadella"}]

custom_output_fields

string

Optional

Output columns to keep, separated by a pipe (|). Use dot notation for nested fields. Omit to return every field. Example: url|name|about.updated_on

discover_by

string

Optional

Discovery method, used with type discover_new. Valid values depend on the scraper; common ones are keyword, best_sellers_url, category_url, location, url and user_name. Example: keyword

include_errors

boolean

Optional

Set to true to include a report of per-input errors (for example dead_page) with the results.

limit_multiple_results

integer

Optional

Cap on the total number of records the job collects across all inputs. Minimum 1. Example: 500

limit_per_input

integer

Optional

Maximum number of records to return for each input. Mostly useful for discovery requests, whose result count is otherwise open-ended. Minimum 1. Example: 10

type

string

Optional

Set to discover_new to run a discovery collection that finds new records (by keyword, category URL, location and so on) instead of collecting the exact inputs. Pair it with discover_by. One of: `discover_new`.

brightdata_trigger_studio_batch

Batch trigger studio

brightdata_unlock_url

Unlock url

For more tools, view docs.

Build your Agent

Same auth pattern across LangChain, OpenAI, Anthropic, and Google ADK.

import { ScalekitClient } from "@scalekit-sdk/node";
import { createReactAgent } from "@langchain/langgraph/prebuilt";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);

// Bright Data tools scoped to this user
const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["brightdata"], toolNames: [
    "brightdata_serp_search",
    "brightdata_unlock_url",
    "brightdata_trigger_scrape"] },
  pageSize: 100,
});

const agent = createReactAgent({ llm, tools });
await agent.invoke({ messages: [{ role: "user", content: "Search Google for MCP gateway vendors and fetch the top result as markdown" }] });

import OpenAI from "openai";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);
const openai = new OpenAI();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["brightdata"] }, pageSize: 100,
});

const res = await openai.chat.completions.create({
  model: "gpt-5",
  messages: [{ role: "user", content: "Search Google for MCP gateway vendors and fetch the top result as markdown" }],
  tools,
});

// Execute the tool call with the user's vaulted Bright Data access token
await sk.tools.executeTool(res.choices[0].message.tool_calls[0], "user_123");

import Anthropic from "@anthropic-ai/sdk";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);
const anthropic = new Anthropic();

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["brightdata"] }, pageSize: 100,
});

const msg = await anthropic.messages.create({
  model: "claude-sonnet-5",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Search Google for MCP gateway vendors and fetch the top result as markdown" }],
  tools,
});

// Tool call runs with the user's vaulted Bright Data access token
await sk.tools.executeTool(msg.content, "user_123");

import { Agent } from "@google/adk/agents";
import { ScalekitClient } from "@scalekit-sdk/node";

const sk = new ScalekitClient(env.SCALEKIT_ENV_URL, env.SCALEKIT_CLIENT_ID, env.SCALEKIT_CLIENT_SECRET);

const { tools } = await sk.tools.listScopedTools("user_123", {
  filter: { connectionNames: ["brightdata"] }, pageSize: 100,
});

const agent = new Agent({
  name: "web_data_agent",
  model: "gemini-2.5-pro",
  instruction: "Bright Data tools scoped to this user",
  tools,
});

await agent.run("Search Google for MCP gateway vendors and fetch the top result as markdown");

Try these prompts

Copy any prompt into your agent. Each maps directly to a Bright Data tool. Click to copy, paste into your agent, done.

Pages and search

Copy the prompt

Copied

Fetch this product page through Web Unlocker and return markdown.

Copy the prompt

Copied

Search Google for 'AI agent auth' and return the organic results.

Copy the prompt

Copied

Show the Web Unlocker success rate for example.com over the past 7 days.

Scrapers

Copy the prompt

Copied

List the pre-built scrapers available on our account.

Copy the prompt

Copied

Trigger a scrape job for these 50 product URLs.

Copy the prompt

Copied

Check the progress of this snapshot and download it when ready.

Datasets and account

Copy the prompt

Copied

Search the Business Search index for fintech companies in Berlin.

Copy the prompt

Copied

List Marketplace datasets and show the fields of this one.

Copy the prompt

Copied

What is our current Bright Data balance?

SEE HOW AUTH WORKS

Each user connects their own Bright Data access token once, and Scalekit sends it with every call. Tokens stay vaulted, every call is scope checked, and every action is logged.

Authorize

Your user connects

Bright Data

once. We tie it to their identity and the meetings they approved — no shared bot account, no org-wide access

Who:

user ‘A’

when:

Once per user

access:

Limited to user

Store

Their

Bright Data

token lives in a vault scoped to them. User A's meetings are never reachable by an agent acting for user B, even on the same connection

vault:

encrypted

scope:

per-user

tokens:

auto-refreshed

Resolve

When your agent calls a

Bright Data

tool, we fetch the right token server-side. It never touches your agent, never appears in the LLM context, never shows up in your logs

speed:

~40ms

check:

before every call

seen by:

nobody

Audit

Every

Bright Data

tool call is logged — who triggered it, which meeting was fetched, what came back. 90 days of history, tied to the user who authorized it

history:

90 days

export:

SIEM-ready

logged:

every call

Test other agents

See the same per-user auth pattern across other search and developer connectors.

GTM and RevOps Teams

Competitive intelligence briefing agent

Scans Gong calls for competitor mentions, matches each one to its Notion battlecard, and DMs every affected rep a single Slack digest per cycle. Every call runs as the PMM who owns the briefing, never a shared bot.

GTM and RevOps Teams

Outbound prospecting agent

Searches Apollo for prospects matching your ICP, scores and ranks them, drafts personalized outreach in Gmail, and logs every send to Google Sheets. Mail goes out as the rep, not from a shared inbox.

Why Scalekit

Secure your agent's access. Connectors ship in minutes

01.

Shared tokens break per-user analytics

A shared Bright Data access token looks fine in a demo. In production every scrape job looks like one service account, and you cannot tell which user triggered it. Scalekit resolves the credential of the actual user who triggered the agent, never a shared bot.

// shared token
audit → bot_service_account

// scalekit
audit → user_abc ✓

02.

Authentication is not authorization

03.

Multi-tenancy is architectural

04.

Bright Data today. Ten connectors tomorrow.

“Our agents act across Salesforce, Gong, Google Drive, and more, on behalf of every customer. Scalekit behind the scenes meant we can keep adding tools without ever rebuilding how credentials or tool calling work.”

Venu Madhav Kattagoni

Head of Engineering / Von

FAQs

Frequently Asked Questions

Does the agent access Bright Data as the user or through a shared token?

As the user. Each user connects their own Bright Data access token once, and Scalekit sends it with every call. Audit logs attribute every scrape job and search to that user, not a shared service account.

Where is the Bright Data access token stored?

In Scalekit's managed AES-256 token vault, namespaced per tenant. To rotate it, the user reconnects with a new access token. Revocation is a single dashboard action. Credentials never appear in prompts, logs, or LLM context.

Can I limit what the agent does in Bright Data?

Yes. Bright Data has no delete tools. Filter by tool name in listScopedTools to leave out tools that start jobs, such as brightdata_trigger_scrape, and brightdata_filter_marketplace_dataset, which creates a billable snapshot. Scalekit also enforces scope checks before every API call.

What happens when a user revokes Bright Data access?

The connection is invalidated on the next tool call. Subsequent requests for that user fail closed with a clear error. Other users in the tenant remain unaffected. The event is logged for audit.

Can the agent check Bright Data spend before a large job?

Yes. brightdata_get_balance returns the account balance, credit, prepayment, and pending costs in USD, and brightdata_get_zone_cost returns billed cost and bandwidth for one zone over a date range.