{"passport":{"unfragile":{"@version":"1.0","version":"2026-05","artifact":{"id":"npm_npm-brightdatamcp","slug":"npm-brightdatamcp","name":"@brightdata/mcp","type":"mcp","url":"https://www.npmjs.com/package/@brightdata/mcp","page_url":"https://unfragile.ai/npm-brightdatamcp","categories":["mcp-servers"],"tags":["mcp","brightdata"],"pricing":{"model":"open_source","free":true,"starting_price":null},"status":"active","verified":false},"capabilities":[{"id":"npm_npm-brightdatamcp__cap_0","uri":"capability://tool.use.integration.web.scraping.via.mcp.protocol","name":"web-scraping-via-mcp-protocol","description":"Exposes Bright Data's web scraping infrastructure through the Model Context Protocol, allowing LLM agents and tools to request structured data extraction from websites without managing proxies, browsers, or anti-bot handling directly. Implements MCP server pattern that translates tool-calling requests into Bright Data API calls, handling authentication, session management, and response parsing transparently.","intents":["I want my AI agent to scrape competitor pricing pages and extract structured product data","I need to fetch real-time data from JavaScript-heavy websites without building browser automation","I want to integrate web scraping into my LLM workflow without managing proxy rotation or CAPTCHA solving"],"best_for":["AI agents and LLM applications requiring live web data","Teams building data pipelines that need LLM-native access to web content","Developers integrating Bright Data into Claude, ChatGPT, or local LLM workflows via MCP"],"limitations":["Requires active Bright Data subscription with API credits for scraping operations","MCP protocol overhead adds latency compared to direct API calls — suitable for agentic workflows, not real-time streaming","Response parsing depends on Bright Data's extraction quality — complex dynamic sites may require custom selectors","No built-in caching — repeated requests to same URL consume credits"],"requires":["Bright Data account with active API credentials","MCP-compatible client (Claude Desktop, LLM frameworks supporting MCP)","Node.js 16+ for running the MCP server","Network access to Bright Data API endpoints"],"input_types":["URL strings","CSS selectors or XPath expressions","JSON configuration objects for scraping parameters"],"output_types":["structured JSON data","HTML/text content","parsed tabular data"],"categories":["tool-use-integration","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-brightdatamcp__cap_1","uri":"capability://tool.use.integration.proxy.rotation.abstraction.for.agents","name":"proxy-rotation-abstraction-for-agents","description":"Abstracts Bright Data's residential and datacenter proxy networks behind MCP tool calls, allowing LLM agents to request data from geographically-specific or IP-rotated contexts without managing proxy pools, authentication, or rotation logic. The MCP server handles proxy selection, failover, and session persistence transparently, presenting a simple semantic interface to the agent.","intents":["I want my agent to fetch data as if it's coming from different geographic regions","I need to bypass IP-based rate limiting or geo-blocking without exposing proxy details to my agent logic","I want to rotate through residential IPs for large-scale data collection without managing proxy infrastructure"],"best_for":["Multi-region data collection agents","Applications requiring geo-spoofing or IP rotation for compliance testing","Teams avoiding infrastructure management for proxy networks"],"limitations":["Proxy rotation adds 500ms-2s latency per request depending on Bright Data network load","Residential proxies are rate-limited by Bright Data's fair-use policy — high-frequency requests may be throttled","No local caching of proxy state — each MCP call re-negotiates proxy selection","Geographic targeting accuracy depends on Bright Data's IP geolocation database"],"requires":["Bright Data account with proxy access tier","MCP client with tool-calling support","API credentials configured in MCP server environment"],"input_types":["URL","geographic region code (e.g., 'US', 'GB', 'JP')","proxy type preference ('residential', 'datacenter', 'isp')"],"output_types":["HTTP response with rotated IP","metadata about proxy used (country, ISP, rotation timestamp)"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-brightdatamcp__cap_2","uri":"capability://tool.use.integration.residential.ip.pool.management.via.mcp","name":"residential-ip-pool-management-via-mcp","description":"Provides MCP tool interface to Bright Data's residential IP pool, enabling agents to request data through specific IP addresses, sticky sessions, or rotating residential proxies without managing pool state locally. Implements session affinity tracking, IP rotation scheduling, and fallback logic server-side, exposing only high-level session and rotation controls to the agent.","intents":["I want my agent to maintain a consistent IP session for multi-step workflows (login, browse, checkout)","I need to rotate through residential IPs at controlled intervals without managing pool state","I want to request data from a specific country's residential IP pool with automatic failover"],"best_for":["Agents performing multi-step workflows requiring session persistence","Large-scale data collection requiring residential IP quality","Applications needing geographic diversity with residential IP authenticity"],"limitations":["Sticky session duration is limited by Bright Data's policy (typically 10-30 minutes) — long-running workflows may require re-authentication","Residential IP availability varies by geography and time — requests may fail if pool is exhausted","No predictive IP rotation — agents cannot pre-select specific IPs, only request rotation","Session state is server-side only — MCP client cannot inspect or manipulate session directly"],"requires":["Bright Data residential proxy subscription","MCP server with persistent session storage (in-memory or external)","API credentials with residential proxy permissions"],"input_types":["session ID or creation request","rotation interval (seconds)","geographic region preference","URL to fetch through session"],"output_types":["session token","current IP address and metadata","HTTP response through residential IP"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-brightdatamcp__cap_3","uri":"capability://tool.use.integration.datacenter.proxy.access.via.mcp","name":"datacenter-proxy-access-via-mcp","description":"Exposes Bright Data's datacenter proxy network through MCP tools, allowing agents to route requests through high-speed datacenter IPs for performance-critical or high-volume data collection. Implements connection pooling, load balancing, and automatic failover server-side, with MCP interface for proxy selection by country, ISP, or performance tier.","intents":["I want my agent to fetch large volumes of data quickly using datacenter proxies","I need to distribute requests across multiple datacenter IPs for parallel data collection","I want to use datacenter proxies for performance testing or load generation without managing proxy infrastructure"],"best_for":["High-volume data collection agents","Performance-critical scraping workflows","Teams needing fast, reliable proxy access without residential IP requirements"],"limitations":["Datacenter IPs are more likely to be blocked by anti-bot systems compared to residential IPs","No session affinity — each request may route through different datacenter IP","Geographic targeting is less granular than residential proxies","Bright Data datacenter pool may be shared, causing potential rate-limiting under high load"],"requires":["Bright Data datacenter proxy subscription","MCP client with tool-calling support","API credentials with datacenter proxy access"],"input_types":["URL","country code for datacenter selection","ISP preference (optional)","concurrency level for parallel requests"],"output_types":["HTTP response through datacenter IP","proxy metadata (country, ISP, response time)"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-brightdatamcp__cap_4","uri":"capability://tool.use.integration.anti.bot.detection.bypass.via.mcp","name":"anti-bot-detection-bypass-via-mcp","description":"Integrates Bright Data's anti-bot detection evasion capabilities (browser fingerprinting, header spoofing, CAPTCHA solving) through MCP tools, allowing agents to request data from protected sites without implementing evasion logic themselves. The MCP server handles browser emulation, header rotation, and CAPTCHA detection/solving transparently, presenting a simple 'fetch URL' interface to the agent.","intents":["I want my agent to scrape sites with anti-bot protection without implementing CAPTCHA solving","I need to fetch data from JavaScript-heavy sites that require browser emulation","I want to bypass Cloudflare, reCAPTCHA, and other anti-bot systems without managing detection logic"],"best_for":["Agents scraping protected or heavily-guarded websites","Applications requiring reliable data extraction from anti-bot-protected sources","Teams avoiding custom anti-bot evasion implementation"],"limitations":["CAPTCHA solving adds 5-30 second latency per request depending on complexity","Browser emulation is slower than direct HTTP requests — suitable for occasional protected-site access, not high-frequency scraping","Anti-bot evasion is an arms race — Bright Data's methods may become ineffective as sites update detection","Some sites explicitly prohibit scraping in ToS — using anti-bot bypass may violate terms"],"requires":["Bright Data account with anti-bot/browser rendering tier","MCP client with tool-calling support","API credentials with anti-bot capabilities enabled"],"input_types":["URL","expected content selector (to verify successful bypass)","timeout for CAPTCHA solving (seconds)"],"output_types":["rendered HTML after anti-bot bypass","extracted structured data","metadata about bypass method used"],"categories":["tool-use-integration","data-processing-analysis"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-brightdatamcp__cap_5","uri":"capability://data.processing.analysis.structured.data.extraction.with.schema.validation","name":"structured-data-extraction-with-schema-validation","description":"Provides MCP tools for requesting structured data extraction from web content using schema-based selectors and validation, allowing agents to specify expected output structure (JSON schema) and receive validated, typed data. Implements server-side extraction using CSS selectors, XPath, or Bright Data's AI-powered extraction, with schema validation and type coercion before returning to agent.","intents":["I want my agent to extract product listings with guaranteed JSON schema structure","I need to validate extracted data matches expected types before processing","I want to extract structured data from unstructured HTML without writing custom parsing logic"],"best_for":["Agents building structured datasets from web scraping","Applications requiring type-safe data extraction","Teams avoiding custom HTML parsing and validation logic"],"limitations":["Schema validation adds latency — complex schemas may require multiple extraction passes","AI-powered extraction (if used) is slower than selector-based extraction but more flexible","Extraction accuracy depends on HTML structure consistency — sites with dynamic layouts may fail validation","No built-in schema inference — agents must define schemas explicitly"],"requires":["Bright Data account with extraction tier","JSON schema definition for expected output","MCP client with tool-calling support"],"input_types":["URL","JSON schema (JSON Schema format)","CSS selectors or XPath expressions","extraction method preference ('selector', 'ai', 'hybrid')"],"output_types":["validated JSON matching schema","extraction metadata (confidence scores, field coverage)","validation error details if schema mismatch"],"categories":["data-processing-analysis","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-brightdatamcp__cap_6","uri":"capability://tool.use.integration.mcp.protocol.server.implementation","name":"mcp-protocol-server-implementation","description":"Implements the Model Context Protocol server specification, exposing Bright Data capabilities as standardized MCP tools that LLM clients (Claude, ChatGPT, local LLMs) can discover and invoke. Uses MCP's JSON-RPC transport layer, tool schema registration, and resource management patterns to bridge Bright Data's REST API into LLM-native tool-calling interfaces.","intents":["I want to use Bright Data with Claude or ChatGPT without custom integrations","I need to expose Bright Data as tools in my LLM framework that supports MCP","I want to build LLM agents that can natively call Bright Data scraping and proxy functions"],"best_for":["Developers building LLM agents with Claude, ChatGPT, or local LLMs","Teams standardizing on MCP for LLM tool integration","Applications requiring LLM-native access to Bright Data without custom API wrappers"],"limitations":["MCP protocol overhead adds ~100-200ms per tool invocation compared to direct API calls","Tool discovery and schema validation happen at client startup — schema changes require client restart","No streaming support for long-running operations — agents must wait for full response","MCP server must be running separately from LLM client — adds deployment complexity"],"requires":["MCP-compatible LLM client (Claude Desktop, LLM framework with MCP support)","Node.js 16+ to run the MCP server","Bright Data API credentials","Network connectivity between LLM client and MCP server"],"input_types":["tool invocation requests (JSON-RPC format)","tool schema queries"],"output_types":["tool execution results (JSON)","tool schema definitions (JSON Schema format)","error responses (JSON-RPC error format)"],"categories":["tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-brightdatamcp__cap_7","uri":"capability://tool.use.integration.authentication.and.credential.management","name":"authentication-and-credential-management","description":"Manages Bright Data API authentication and credential lifecycle through the MCP server, handling token refresh, credential rotation, and secure storage of API keys. Implements credential injection into Bright Data API calls without exposing keys to LLM clients, using environment variables or secure config files for credential storage.","intents":["I want to use Bright Data with my LLM agent without exposing API keys to the agent","I need to rotate Bright Data credentials without restarting the MCP server","I want to manage multiple Bright Data accounts through a single MCP server"],"best_for":["Production LLM agent deployments requiring secure credential handling","Teams managing multiple Bright Data accounts","Applications requiring credential rotation without downtime"],"limitations":["Credentials must be stored in environment or config files — no built-in secrets management (e.g., HashiCorp Vault integration)","Token refresh logic is synchronous — expired tokens may cause request failures if refresh is slow","No audit logging of credential usage — cannot track which agents used which credentials","Credential rotation requires MCP server restart or manual refresh trigger"],"requires":["Bright Data API credentials (API key or OAuth token)","Environment variable or config file for credential storage","MCP server with credential management module"],"input_types":["API key or OAuth token","credential rotation trigger"],"output_types":["authenticated API requests to Bright Data","credential status (valid, expired, invalid)"],"categories":["tool-use-integration","safety-moderation"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-brightdatamcp__cap_8","uri":"capability://tool.use.integration.error.handling.and.retry.logic","name":"error-handling-and-retry-logic","description":"Implements server-side error handling and automatic retry logic for Bright Data API calls, including exponential backoff, circuit breaker patterns, and graceful degradation. Translates Bright Data API errors into MCP-compatible error responses, allowing agents to understand failure reasons without implementing retry logic themselves.","intents":["I want my agent to automatically retry failed scraping requests without custom retry code","I need to understand why a Bright Data request failed and get actionable error details","I want to prevent cascading failures when Bright Data API is temporarily unavailable"],"best_for":["Production agents requiring resilience to transient failures","Applications with strict uptime requirements","Teams avoiding custom retry and error-handling logic"],"limitations":["Retry logic adds latency — exponential backoff may delay responses by 10-60 seconds on repeated failures","Circuit breaker may reject requests during Bright Data outages — agents cannot override circuit breaker","Error details are limited to Bright Data API responses — underlying network errors may be opaque","No adaptive retry strategy — retry count and backoff are fixed, not tuned per request type"],"requires":["MCP server with retry and error-handling module","Bright Data API credentials"],"input_types":["Bright Data API request","retry configuration (max retries, backoff strategy)"],"output_types":["successful response or final error after retries exhausted","error details (error code, message, retry count)"],"categories":["tool-use-integration","automation-workflow"],"confidence":0.5,"matches":0,"success_rate":0},{"id":"npm_npm-brightdatamcp__cap_9","uri":"capability://automation.workflow.request.rate.limiting.and.quota.management","name":"request-rate-limiting-and-quota-management","description":"Enforces rate limiting and quota management for Bright Data API calls through the MCP server, preventing agents from exceeding Bright Data's rate limits or account quotas. Implements token bucket rate limiting, quota tracking per agent/session, and graceful degradation when limits are approached, with MCP tools for checking remaining quota.","intents":["I want to prevent my agent from exceeding Bright Data rate limits","I need to track and enforce per-agent quotas for Bright Data usage","I want to know how much Bright Data quota remains before making requests"],"best_for":["Multi-agent deployments sharing a single Bright Data account","Applications with strict budget constraints on Bright Data usage","Teams requiring quota visibility and enforcement"],"limitations":["Rate limiting adds latency — requests may be queued if rate limit is approached","Quota tracking is server-side only — no distributed quota enforcement across multiple MCP servers","No predictive quota management — agents cannot reserve quota in advance","Rate limit configuration is static — cannot adjust limits without server restart"],"requires":["MCP server with rate limiting and quota management module","Bright Data account with known rate limits and quotas"],"input_types":["rate limit configuration (requests per second, quota per agent)","quota check request"],"output_types":["rate-limited API response or queue status","remaining quota and rate limit status"],"categories":["automation-workflow","tool-use-integration"],"confidence":0.5,"matches":0,"success_rate":0}],"trust":{"score":26,"verified":false,"data_access_risk":"high","permissions":["Bright Data account with active API credentials","MCP-compatible client (Claude Desktop, LLM frameworks supporting MCP)","Node.js 16+ for running the MCP server","Network access to Bright Data API endpoints","Bright Data account with proxy access tier","MCP client with tool-calling support","API credentials configured in MCP server environment","Bright Data residential proxy subscription","MCP server with persistent session storage (in-memory or external)","API credentials with residential proxy permissions"],"failure_modes":["Requires active Bright Data subscription with API credits for scraping operations","MCP protocol overhead adds latency compared to direct API calls — suitable for agentic workflows, not real-time streaming","Response parsing depends on Bright Data's extraction quality — complex dynamic sites may require custom selectors","No built-in caching — repeated requests to same URL consume credits","Proxy rotation adds 500ms-2s latency per request depending on Bright Data network load","Residential proxies are rate-limited by Bright Data's fair-use policy — high-frequency requests may be throttled","No local caching of proxy state — each MCP call re-negotiates proxy selection","Geographic targeting accuracy depends on Bright Data's IP geolocation database","Sticky session duration is limited by Bright Data's policy (typically 10-30 minutes) — long-running workflows may require re-authentication","Residential IP availability varies by geography and time — requests may fail if pool is exhausted","builder identity is not verified yet","no observed match outcomes yet"],"rank_breakdown":{"adoption":0.05,"quality":0.3,"ecosystem":0.36,"match_graph":0.25,"freshness":0.52,"weights":{"adoption":0.25,"quality":0.25,"ecosystem":0.15,"match_graph":0.23,"freshness":0.12}},"observed_outcomes":{"matches":0,"success_rate":0,"avg_confidence":0,"top_intents":[],"last_matched_at":null},"maintenance":{"status":"active","updated_at":"2026-05-24T12:16:23.903Z","last_scraped_at":"2026-05-03T14:23:54.399Z","last_commit":null},"community":{"stars":null,"forks":null,"weekly_downloads":null,"model_downloads":null,"model_likes":null}},"distribution":{"claim_url":"https://unfragile.ai/submit?claim=npm-brightdatamcp","compare_url":"https://unfragile.ai/compare?artifact=npm-brightdatamcp"}},"signature":"PgiNLdWkbiGmQKAKpVvguJjkpM22mZNqn1Cl+K6QDlCZjHV7uhs0ESnSGB4ZedT4s5S6/UEXjoqwPoZrAzJSCQ==","signedAt":"2026-06-19T21:51:24.963Z","signedBy":"unfragile.ai","version":1},"_links":{"self":"https://unfragile.ai/api/v1/passport/npm-brightdatamcp","artifact":"https://unfragile.ai/npm-brightdatamcp","verify":"https://unfragile.ai/api/v1/verify?slug=npm-brightdatamcp","publicKey":"https://unfragile.ai/api/v1/trust-passport-public-key","spec":"https://unfragile.ai/trust","schema":"https://unfragile.ai/schema.json","docs":"https://unfragile.ai/docs"}}