🚀 Nexus-IQ: Enterprise-Grade Competitive Intelligence Actor

**Nexus-IQ** is a high-performance, real-time competitive intelligence engine designed for GTM teams, Product Managers, and AI Agents. It orchestrates complex extraction, normalization, and AI-driven analysis across **16+ premium B2B platforms and communities** simultaneously.

Built natively with Model Context Protocol (MCP) support, Nexus-IQ empowers autonomous AI agents to research competitors with surgical precision while bypassing anti-bot measures and delivering token-optimized insights.

---

🌟 Key Capabilities

• 🌐 16+ Premium Data Sources: Unified access to G2, Capterra, Reddit, LinkedIn, Twitter/X, Glassdoor, and niche App Stores.
• 🛡️ CloakBrowser™ Stealth: Uses custom C++ source-level Chromium patches to bypass Cloudflare Turnstile, Datadome, Akamai, and CAPTCHA checks without runtime script leakage.
• 🤖 Multi-LLM Intelligence: Unified summaries and structured market signals via Gemini, OpenAI, or Anthropic.
• 🔌 MCP Native: Built-in support to run as an MCP Tool for Claude Desktop, Cursor, Claude Code, and other agentic environments.
• 💰 Smart Proxy Guardrails: Automated Residential Proxy budget enforcement and memory monitors to prevent runtime cost overrun.
• ⚡ Direct Browser DOM Extraction: Extracts the fully-rendered HTML/JSON directly from active browser pages, preventing WAF cookie-binding mismatches.
• 🌍 Multi-lingual Summaries: Delivers competitive reports in English, Indonesian, Spanish, German, or French.

---

📊 Supported Intelligence Sources

Tier 1 (Anti-Bot Heavy)	Tier 2 (Social & Community)	Tier 3 (App Stores & Niche)
🔹 G2	💬 Reddit	📱 Google Play Store
🔹 Capterra	💬 Trustpilot	📱 Apple App Store
🔹 TrustRadius	💬 Product Hunt	🏢 Crunchbase
🔹 Glassdoor	💬 Hacker News	🏢 PeerSpot
🔹 LinkedIn Jobs	💬 Twitter / X	🏢 Software Advice

---

🧠 Technical Architecture

graph TD
    User([User / AI Agent]) -->|Input Schema| Engine[Nexus-IQ Engine]
    Engine -->|1. Proxy Config| ProxyConfig[Apify Proxy Manager]
    ProxyConfig -->|Allocates| ResidentialProxy[Residential/Premium Proxies]
    Engine -->|2. Batch Dispatch| Batcher[Batcher Manager]
    
    Batcher -->|Tier 1 & 2: Stealth| CloakBrowser[CloakBrowser Chromium]
    CloakBrowser -->|Direct DOM Capture| HTMLOutput[Rendered HTML / JSON]
    
    Batcher -->|Tier 3: Standard| gotScraping[got-scraping client]
    gotScraping -->|Fetch check| PlaceholderHTML[Clean Fallback Data]
    
    HTMLOutput & PlaceholderHTML -->|Normalize| RawReviews[(Raw Reviews Datasets)]
    
    RawReviews -->|3. AI Tier| AIProvider[AI Provider Factory]
    AIProvider -->|Summarize Source| Summary[Source Summaries]
    Summary -->|Global Analysis| GlobalInsight[Global Insight Generator]
    
    GlobalInsight & RawReviews -->|4. Final Output| Output([Combined JSON Output])

---

⚙️ Configuration (Input Schema)

Configured via the standard Apify interface or JSON input:

Parameter	Type	Description	Default
target_company *	String	The competitor or brand name to research.	-
sources *	Array	Select from 16 available intelligence providers.	["G2", "Capterra", "Trustpilot"]
timeframe_days	Integer	Data depth in days.	7
target_language	Enum	Output language (english, indonesia, spanish, german, french).	english
ai_provider	Enum	AI summarizer (gemini, openai, anthropic).	gemini
ai_model	String	Specific model identifier (e.g., gemini-1.5-flash, gpt-4o-mini).	gemini-1.5-flash
proxyConfiguration	Object	Standard Apify Proxy Settings. Residential/Premium proxies are recommended.	{ "useApifyProxy": true }
max_proxy_cost_usd	Float	Hard limit for proxy expenditure guardrail.	5.0
skip_ai	Boolean	Return raw reviews directly without AI summarization.	false
force_refresh	Boolean	Bypass cache and force fresh crawling.	false

* Required fields

---

🔌 Model Context Protocol (MCP) Integration

Because Nexus-IQ is hosted directly on the Apify platform, AI agents can interact with it serverless-ly. Below are the configurations to integrate this Actor into various popular AI clients:

1. 💬 Claude Desktop

Add the configuration to your claude_desktop_config.json (typically located at %APPDATA%\Claude\claude_desktop_config.json on Windows or ~/Library/Application Support/Claude/claude_desktop_config.json on macOS):

{
  "mcpServers": {
    "nexus-iq": {
      "command": "npx",
      "args": [
        "-y",
        "@apify/actors-mcp-server",
        "--tools",
        "rexreus/Nexus-IQ"
      ],
      "env": {
        "APIFY_TOKEN": "YOUR_APIFY_API_TOKEN"
      }
    }
  }
}

---

2. 🎛️ Cursor

Cursor supports MCP servers directly through its graphical settings interface:

1. Go to Settings (Gear Icon) -> Features -> MCP.
2. Click + Add New MCP Server.
3. Configure the server with these parameters:
   • Name: nexus-iq
   • Type: command
   • Command: npx -y @apify/actors-mcp-server --tools rexreus/Nexus-IQ
4. Under Environment Variables, add:
   • Key: APIFY_TOKEN
   • Value: YOUR_APIFY_API_TOKEN
5. Click Save.

---

3. 🌊 Windsurf

Add the configuration to your Windsurf global mcp_config.json (typically located at ~/.codeium/windsurf/mcp_config.json):

{
  "mcpServers": {
    "nexus-iq": {
      "command": "npx",
      "args": [
        "-y",
        "@apify/actors-mcp-server",
        "--tools",
        "rexreus/Nexus-IQ"
      ],
      "env": {
        "APIFY_TOKEN": "YOUR_APIFY_API_TOKEN"
      }
    }
  }
}

---

4. 🤖 Cline (VS Code Extension) / Hermes

Cline uses a dedicated configuration file named cline_mcp_settings.json (typically located in %APPDATA%\Code\User\globalStorage\saoudrizwan.claude-dev\settings\cline_mcp_settings.json on Windows):

{
  "mcpServers": {
    "nexus-iq": {
      "command": "npx",
      "args": [
        "-y",
        "@apify/actors-mcp-server",
        "--tools",
        "rexreus/Nexus-IQ"
      ],
      "env": {
        "APIFY_TOKEN": "YOUR_APIFY_API_TOKEN"
      },
      "disabled": false
    }
  }
}

---

5. 💻 Claude Code (CLI)

You can add the tool to Claude Code by running this command in your terminal:

$claude mcp add nexus-iq npx -y @apify/actors-mcp-server --tools rexreus/Nexus-IQ

Make sure the APIFY_TOKEN environment variable is exported in your shell session (export APIFY_TOKEN=...) so that Claude Code inherits it.

---

🛠️ Exposed MCP Tools (Skills)

Once the server is configured, the following tool (or "skill") is exposed to your AI agent:

⚙️ nexus_iq_get_sentiment

Description: Triggers the Apify Actor to gather competitive intelligence and execute sentiment analysis on the target company.

Tool Parameters:

Parameter	Type	Required	Description	Default
target_company	string	Yes	The competitor/brand name to research.	-
sources	array	Yes	Array of target platforms to scrape. Choose from: G2, Capterra, TrustRadius, Reddit, Glassdoor, Trustpilot, Product Hunt, Hacker News, LinkedIn Jobs, Twitter/X, GetApp, Software Advice, PeerSpot, Crunchbase, Google Play, Apple App Store.	-
timeframe_days	integer	No	Number of days in the past to pull reviews.	7
target_language	string	No	Output language for AI insights.	"english"
ai_provider	string	No	Preferred AI provider (gemini, openai, anthropic).	"gemini"
ai_model	string	No	Specific model identifier (e.g. gemini-1.5-flash).	-
skip_ai	boolean	No	Return raw JSON data without AI summarization.	false

---

🤖 Example Agent Tasks (How to Prompt your AI Agent)

Once the MCP integration is complete, you can give your AI agent complex commands directly. Here are sample prompts you can copy and use:

[!TIP] Task 1: Competitor Feature & Sentiment Gap "Use the Nexus-IQ tool to research G2, Reddit, and Twitter reviews for competitor 'Apify' over the last 14 days. Summarize the results in Indonesian, focusing on the top complaints and feature requests."

[!TIP] Task 2: Churn & Pricing Signal Check "Identify if there are any churn risks or pricing concerns for 'Salesforce' by running a Nexus-IQ query on G2 and Capterra. Compare what you find with HubSpot."

[!TIP] Task 3: App Store Sentiment Audit "Audit user reviews for 'Slack' on Google Play and Apple App Store using Nexus-IQ. Extract the raw reviews and compile a summary of competitor mentions in German."

---

📦 Output Schema

Nexus-IQ returns a structured JSON output that contains metadata, clean review text datasets, and standard market intelligence schemas:

{
  "query_metadata": {
    "target_company": "Apify",
    "timeframe_days": 7,
    "sources_requested": ["G2", "Google Play"],
    "sources_successful": ["G2", "Google Play"],
    "total_reviews": 32,
    "timestamp": "2026-06-05T08:00:00.000Z",
    "latency_ms": 28400,
    "ai_provider": "gemini",
    "ai_model": "gemini-1.5-flash"
  },
  "global_insight": {
    "summary": "Users praise the extensive developer capabilities and proxy integration, but cite complex javascript selector setup as a major learning curve.",
    "sentiment": {
      "average_rating": 4.5,
      "total_reviews": 32,
      "sentiment_score": 0.85,
      "top_praises": ["stealth capabilities", "developer experience", "integrations"],
      "top_complaints": ["learning curve", "datacenter IP blocking"]
    },
    "signals": {
      "churn_risk": false,
      "feature_requests": ["No-code selector builder"],
      "pricing_mentions": ["granularity of computation tiers"],
      "competitor_mentions": ["BrightData", "ZenRows"]
    }
  },
  "raw_reviews": [
    {
      "id": "a9f8b7c6...",
      "source": "G2",
      "author": "Steve D.",
      "rating": 5,
      "title": "Standard for web scaling",
      "text": "Apify actors make it easy to scale web extraction tasks safely.",
      "date": "2026-06-03T12:00:00.000Z",
      "url": "https://www.g2.com/products/apify/reviews"
    }
  ]
}

---

🔒 Security & Privacy

• BYOK (Bring Your Own Key): Provider API keys are handled as secrets and are never stored on Apify's datasets or printed in the log.
• PII Masking: Automatic scrubbing of sensitive client data (e.g. user emails, phone numbers) before passing it to AI analysis.
• Challenge Isolation: If a source is blocked by an unresolvable anti-bot challenge, it is cleanly reported under failed_sources rather than returning mock/corrupted data.

---

Developed by the Nexus-IQ Engineering Team. Optimized for Autonomous AI.