novada-proxy-workspace

Name: novada-proxy-workspace
Author: NovadaLabs

data-ai MCP Server

Residential proxy MCP server for AI agents — bypass Cloudflare/Akamai, geo-target 195+ countries, sticky sessions, Google search. Works with Claude, Cursor, Windsurf. Multi-provider: Novada, BrightData, Smartproxy, Oxylabs.

byNovadaLabs

VerifiedInstall Ready

data-aidata-ai gcp

5 views4 stars1 forksMIT

View on GitHub Homepage Docs

Why This Matters

Discovered via github-topic:mcp-server and last synced 3mo ago.

VerifiedInstall Ready

Source: github-topic:mcp-server
Stars: 4
Last synced: 3mo ago
Install: Instructions detected

Install

1. Install the package

npm install novada-proxy-workspace

Tools

Resources

Prompts

Standard I/O

Transport

Available Tools (67)

novada_proxy_map

Crawl a URL and return all internal links as JSON array

Description

Key Arguments

fields

string[]

include_content

boolean

TLS_ERROR

TLS/SSL 连接失败

UNKNOWN_ERROR

Unexpected error

ISP

6-hour sticky, stable IPs

Limitation

Workaround

credits_estimated

每个响应都内置成本追踪

depth

string

novada_proxy_session

Sticky session — same IP across every call

research_topic

搜索 + 批量阅读工作流 — 搜索主题并阅读排名靠前的页面

url

string

concurrency

number

Behavior

Detail

wait_for

string

NETWORK_ERROR

DNS 解析失败

Countries

195+

extract_product

从任意电商 URL 提取结构化产品数据

Tool

What It Does

novada_proxy_render

Render JS-heavy pages with real Chromium [BETA]

crawl_site

发现站点所有页面，然后并行抓取

city

string

num

number

TIMEOUT

请求超时

Client

Install method

Datacenter

Lowest latency, cheapest per GB

limit

number

Problem

Solution

novada_proxy_search

Google search -> structured JSON (title, url, snippet)

novada_proxy_batch_fetch

Fetch 2–20 URLs concurrently (up to 5x parallel)

Feature

What It Means

BOT_DETECTION_SUSPECTED

被目标站点封锁（403）

language

string

PROVIDER_NOT_CONFIGURED

缺少凭证

Residential

Real home IPs, hardest to detect

Any

All types handle most sites

novada_proxy_fetch

Fetch any URL through residential proxy

novada_proxy_research

One-shot deep research — search + fetch + synthesize

troubleshoot

代理故障逐步诊断

session_id

string

RATE_LIMITED

HTTP 429 限速

Metric

Value

7 BrightData

`BRIGHTDATA_USER` + `BRIGHTDATA_PASS`

novada_proxy_extract

Extract structured fields — heuristic mode (fields) or LLM mode (schema)

Smartproxy

`SMARTPROXY_USER` + `SMARTPROXY_PASS`

timeout

number

include_external

boolean

verify_sticky

boolean

novada_proxy_crawl

Recursively crawl a site (BFS, depth 1-5) with URL discovery

fetch_url

通过住宅代理抓取 URL，自动绕过反机器人

Parameter

Type

urls

string[]

render_fallback

boolean

Code

Meaning

INVALID_INPUT

参数错误

Tests

Pass

query

string

Disable

Set `PROXY4AGENT_CACHE_TTL_SECONDS=0`

SESSION_STICKINESS_FAILED

Same IP not maintained

Uptime

99.9%

Oxylabs

`OXYLABS_USER` + `OXYLABS_PASS`

format

string

schema

object

novada_proxy_status

## 工具选择决策树 ``` 目标：抓取单个 URL ├─ 静态 HTML 页面？ → novada_proxy_fetch ├─ 需要特定字段（价格/标题）？ → novada_proxy_extract（fields 或 schema 模式） └─ React/Vue SPA / 内容为空？ → novada_proxy_render 目标：批量抓取多个 URL ├─ 已有 URL 列表？ → novada_proxy_batch_fetch ├─ 需要获取单页链接？ → novada_proxy_map → novada_proxy_batch_fetch └─ 需要爬取整个站点？ → novada_proxy_crawl → novada_proxy_batch_fetch 目标：研究某个主题 → novada_proxy_research（一次调用搜索 + 抓取 + 分析）目标：网页搜索 → novada_proxy_search → novada_proxy_batch_fetch 目标：登录 + 多步骤流程 → novada_proxy_session（相同 session_id）目标：验证代理是否正常工作 → novada_proxy_status ``` --- ## 三大流水线模式 ### 全站爬取流水线（推荐） ``` novada_proxy_map(url, limit=50) │ ▼ 返回 20-50 个内部链接（1 credit，4 秒） │ ▼ novada_proxy_batch_fetch(urls, concurrency=5) │ ▼ 并发抓取所有页面（N credits，~4 秒，比串行快 15x） ``` ### 搜索研究流水线 ``` novada_proxy_search(query, num=10) │ ▼ 结构化 JSON：标题 + URL + 摘要 │ ▼ novada_proxy_batch_fetch(urls, format="markdown") │ ▼ 全部页面内容并行返回 ``` ### 粘性会话 — 登录 + 多页抓取 ``` session_id = "job_001" novada_proxy_session(session_id, url="/login") → 同一 IP novada_proxy_session(session_id, url="/dashboard") → 同一 IP novada_proxy_session(session_id, url="/data/1") → 同一 IP novada_proxy_session(session_id, url="/data/2") → 同一 IP ``` --- ## 智能体优先设计 > [!NOTE] > Novada Proxy 是唯一专为自主 AI 智能体设计的代理 MCP。每个响应、错误和描述都为机器消费而优化。

country

string

Works Well With - Skills

implementing-gcp-vpc-firewall-rules

Implementing and auditing GCP VPC firewall rules to enforce network segmentation, restrict ingress and egress traffic, apply hierarchical firewall policies across the organization, and monitor firewall rule effectiveness using VPC Flow Logs.

61 starsHas guide

auditing-gcp-iam-permissions

Auditing Google Cloud Platform IAM permissions to identify overly permissive bindings, primitive role usage, service account key proliferation, and cross-project access risks using gcloud CLI, Policy Analyzer, and IAM Recommender.

61 starsHas guide

configuring-identity-aware-proxy-with-google-iap

Configuring Google Cloud Identity-Aware Proxy (IAP) to enforce per-request identity verification for Compute Engine, App Engine, Cloud Run, and GKE services using access levels, context-aware policies, and programmatic access with service accounts.

61 starsHas guide

implementing-gcp-binary-authorization

Implement GCP Binary Authorization to enforce deploy-time security controls that ensure only trusted, attested container images are deployed to Google Kubernetes Engine and Cloud Run.

61 starsHas guide

Works Well With - Agents

Generative Ai

Sample code and notebooks for Generative AI on Google Cloud, with Gemini Enterprise Agent Platform

17757 stars

Warden

The secure gateway connecting AI agents to enterprise systems.

166 stars

Decisionbox Platform

DecisionBox connects to your data warehouse, runs autonomous AI agents that write and execute SQL, and surfaces validated insights and actionable recommendations — without you asking a single question.

91 stars