mcp-open-webresearch

Version Issues

Proxy-aware Model Context Protocol (MCP) server for web searching and content extraction.

Designed to be robust and compatible with various network environments, including those using SOCKS and HTTP proxies.

Features

Dynamic Engine Discovery: Engines are loaded dynamically from the src/infrastructure/search/ directory. Adding a new engine requires only a new folder and file, without modifying core logic.
Multi-Engine Search: Aggregates results from Bing, DuckDuckGo, and Brave.
Deep Research (search_deep): Recursive research agent that performs multi-round searching, citation extraction, and answer synthesis.
Ephemeral Downloads: In-memory storage for Deep Search reports using a 100MB bounded LRU cache with 10-minute auto-expiration.
Centralized Throttling: Rate limit management (search and pagination cooldowns) across prioritized engines.
Smart Fetch: Configurable fetching utility (impit) with two operational profiles:
- Browser Mode: Includes modern browser headers (User-Agent, Client Hints) for compatibility with sites requiring browser-standard requests.
- Standard Mode: Uses a minimal HTTP client profile for environments where browser-like identification is not required.
Result Sampling: Optional LLM-based filtering to assess result relevance.
Content Extraction: Webpage visiting and markdown extraction tool (visit_webpage) using a headless browser.
Proxy Support: Full support for SOCKS5, HTTPS, and HTTP proxies.
Configuration: Configurable via environment variables and CLI arguments.
Deployment: Docker images available for production and testing.

Credits

This project includes work from the following contributors:

Manav Kundra: Initial implementation of the server.
Aasee: Added multiple search engines and Docker support.
mzxrai: Core logic for the visit_page tool.

Installation & Quick Start

Docker (Recommended)

Latest Stable Release:

docker pull ghcr.io/rinaldowouterson/mcp-open-webresearch:latest
docker run -p 3000:3000 ghcr.io/rinaldowouterson/mcp-open-webresearch:latest

Test/Debug Image:

docker pull ghcr.io/rinaldowouterson/mcp-open-webresearch:test

Local Installation

To run the server locally (e.g., in Claude Desktop or Cline):

[!NOTE] Replace /absolute/path/to/project with your actual project path.

Configuration (mcp_config.json):

{
  "mcpServers": {
    "open-webresearch": {
      "command": "npm",
      "args": [
        "run",
        "start:sampling",
        "--silent",
        "--prefix",
        "/absolute/path/to/project"
      ],
      "headers": {},
      "disabled": false
    }
  }
}

Remote Server (Streamable HTTP)

Endpoint: http://localhost:3000/mcp

Configuration:

{
  "mcpServers": {
    "open-webresearch": {
      "serverUrl": "http://localhost:3000/mcp",
      "headers": {}
    }
  }
}

Client Configuration & Timeouts

Deep Search processes can take several minutes to complete. Some MCP clients (like Cline and RooCode) have a default timeout of 60 seconds, which will cause the operation to fail.

You MUST configure a higher timeout in your client settings.

Cline (`cline_mcp_settings.json`)

Add the "timeout" parameter (in seconds). Recommended: 1800 (30 minutes).

{
  "mcpServers": {
    "open-webresearch": {
      "disabled": false,
      "timeout": 1800,
      "type": "stdio",
      "command": "npm",
      "args": [
        "run",
        "start:sampling",
        "--silent",
        "--prefix",
        "/absolute/path/to/mcp-open-webresearch"
      ],
      "autoApprove": []
    }
  }
}

RooCode (`mcp_settings.json`)

RooCode also respects the timeout parameter.

{
  "mcpServers": {
    "open-webresearch": {
      "disabled": false,
      "timeout": 1800,
      "command": "npm",
      "args": [
        "run",
        "start:sampling",
        "--silent",
        "--prefix",
        "/absolute/path/to/mcp-open-webresearch"
      ],
      "alwaysAllow": []
    }
  }
}

Antigravity / Windsurf (`mcp_config.json`)

Antigravity / Windsurf handles long-running tools natively, but if they let you configure a timeout, it is best practice to do so.

{
  "mcpServers": {
    "open-webresearch": {
      "command": "npm",
      "args": [
        "run",
        "start:sampling",
        "--silent",
        "--prefix",
        "/absolute/path/to/mcp-open-webresearch"
      ],
      "disabled": false
    }
  }
}

Developer Guide: Adding New Engines

To add a new search engine:

Create Directory: src/infrastructure/search/{engine_name}/
Implement Logic: Create {engine_name}.ts with the fetching/parsing logic.

Export Interface: Create index.ts exporting the SearchEngine interface:

import type { SearchEngine } from "../../../types/search.js";
import { searchMyEngine } from "./my_engine.js";
import { isThrottled } from "../../throttle.js"; // Optional

export const engine: SearchEngine = {
  name: "my_engine",
  search: searchMyEngine,
  isRateLimited: () => isThrottled("my_engine"),
};

Restart: The server will automatically discover and load the new engine.

Build and Run

Locally

# 1. Clone
git clone https://github.com/rinaldowouterson/mcp-open-webresearch.git
cd mcp-open-webresearch

# 2. Install
npm install

# 3. Build & Start
npm run build
npm start

Docker

# Production
docker build -t mcp-websearch .
docker run -p 3000:3000 mcp-websearch

# Testing
npm run test:docker

Testing

Unit & E2E Tests

Uses Vitest for testing. Includes dynamic contract tests for all discovered engines.

npm test

Compliance Tests

Verifies the "Smart Fetch" behavior (User-Agent headers) usage using a local mock server.

npm run test .test/engines/smart_fetch_mode.test.ts

Infrastructure Validation

Validates Docker image builds and basic functionality.

npm run test:infrastructure

Available Scripts

Command	Description
`npm run build`	Compiles TypeScript to `build/` folder.
`npm run watch`	Recompiles on file changes.
`npm run inspector`	Launches MCP inspector UI.
`npm start`	Runs the compiled server.
`npm test`	Runs local tests.
`npm run test:docker`	Runs tests in Docker container.
`npm run test:infrastructure`	Validates docker images.
`npm run generate-certs`	Generates self-signed certificates for testing.

Configuration

Configuration is managed via Environment Variables or CLI arguments.

Variable	Default	Description
`PORT`	`3000`	Server port.
`PUBLIC_URL`	`http://localhost:port`	Public URL for download links.
`ENABLE_CORS`	`false`	Enable CORS.
`CORS_ORIGIN`	`*`	Allowed CORS origin.
`DEFAULT_SEARCH_ENGINES`	`bing,duckduckgo,brave`	Default engines list.
`ENABLE_PROXY`	`false`	Enable proxy support.
`HTTP_PROXY`	-	HTTP Proxy URL.
`HTTPS_PROXY`	-	HTTPS Proxy URL.
`SOCKS5_PROXY`	-	SOCKS5 Proxy URL (Highest Priority).
`SAMPLING`	`false`	Enable result sampling.
`SKIP_IDE_SAMPLING`	`false`	Prefer external API over IDE.
`LLM_BASE_URL`	-	External LLM API base URL.
`LLM_API_KEY`	-	External LLM API key.
`LLM_NAME`	-	External LLM model name.
`LLM_TIMEOUT_MS`	`30000`	Timeout for external LLM calls.
`DEEP_SEARCH_MAX_LOOPS`	`20`	Max research iterations.
`DEEP_SEARCH_RESULTS_PER_ENGINE`	`5`	Results per engine per round.
`DEEP_SEARCH_SATURATION_THRESHOLD`	`0.6`	Threshold to stop research early.
`DEEP_SEARCH_MAX_CITATION_URLS`	`10`	Max URLs to visit for citations.
`DEEP_SEARCH_REPORT_RETENTION_MINUTES`	`10`	Download expiration time (minutes).
`WRITE_DEBUG_TERMINAL`	`false`	Log debug output to stdout.
`WRITE_DEBUG_FILE`	`false`	Log debug output to file.

CLI Arguments

CLI arguments override environment variables.

Argument	Description
`--port <number>`	Port to listen on.
`--debug`	Enable debug logging (stdout).
`--debug-file`	Enable debug logging (file).
`--cors`	Enable CORS.
`--proxy <url>`	Proxy URL (http, https, socks5).
`--engines <items>`	Comma-separated list of engines.
`--sampling`	Enable sampling.
`--no-sampling`	Disable sampling.

Search Pipeline & Scoring

The server uses a multi-stage pipeline to aggregate and refine search results:

1. Multi-Engine Retrieval

Concurrent requests are dispatched to all configured engines (Bing, Brave, DuckDuckGo). Raw results are collected into a single pool.

2. Consensus Scoring & Deduplication

Results are grouped by their canonical URL (protocol/www-agnostic hash).

Deduplication: Multiple entries for the same URL are merged.
Scoring: A consensusScore is calculated for each unique URL:
- Inverted Rank Sum: Sum of inverted ranks ($1/rank$) across engines. Higher placement results in a higher score.
- Engine Boost: Multiplies the sum by the number of unique engines that identified the URL. This prioritizes multi-provider agreement.
Sorting: The final list is sorted by the calculated consensusScore in descending order.

3. LLM Sampling (Optional)

If SAMPLING=true, the top-ranked results are sent to an LLM to evaluate semantic relevance to the query.

Filtering: Sampling acts as a binary filter. It removes results identified as irrelevant (spam, off-topic).
Final Set: The original consensus scores are preserved. Only the composition of the list changes.

LLM Sampling Strategy

When sampling is enabled, the server follows a tiered resolution logic to select which LLM to use:

SKIP_IDE_SAMPLING	IDE Available	API Configured	Resolution
`false` (default)	✅	✅	IDE Sampling
`true`	✅	❌	IDE Sampling
`false`	❌	✅	External API
`true`	✅ OR ❌	✅	External API
`false` OR `true`	❌	❌	No Sampling

[!TIP] You can use a model without API key, the LLM_API_KEY value is optional.

[!IMPORTANT] Deep Search Compatibility: The search_deep tool strictly requires LLM capability (either via IDE or API). If neither is available, the tool will appear in the MCP list but will throw an error upon execution.

Tools Documentation

`search_deep`

Recursive research agent for deep investigation. Searches multiple sources, extracts citations, and synthesizes a comprehensive answer.

Requires LLM Sampling capability.

Input:

{
  "objective": "Deep research goal",
  "max_loops": 3,
  "results_per_engine": 5,
  "max_citation_urls": 10,
  "engines": ["bing", "brave"],
  "attach_context": false
}

Output: A structured Markdown report including a reference list. If configured, a Download URL at the top of the output permits downloading the results as a file.

`search_web`

Performs a search across configured engines.

Input:

{
  "query": "search query",
  "max_results": 10,
  "engines": ["bing", "brave"],
  "sampling": true
}

`visit_webpage`

Visits a URL and returns markdown content.

Input:

{
  "url": "https://example.com/article",
  "capture_screenshot": false
}

`set_engines`

Updates default search engines.

Input:

{
  "engines": ["duckduckgo", "brave"]
}

`get_engines`

Returns configured search engines.

`set_sampling`

Enables or disables result sampling.

Input:

{
  "enabled": true
}

`get_sampling`

Returns current sampling status.

📥 Ephemeral Downloads

Deep Search results are served via an in-memory buffer cache.

Storage: Reports are stored as Buffer objects in the C++ heap to avoid V8 string memory limits.
Expiration: Each individual entry expires exactly 10 minutes after creation. Access operations (get) do not extend the time-to-live (TTL).
Memory Safety: The cache is bounded by a 100MB ceiling. When the limit is reached, a Least Recently Used (LRU) eviction policy removes the oldest entries.
URL Configuration: Link generation depends on the PUBLIC_URL variable to ensure accessible download endpoints in proxied environments.

Roadmap

Deep Search: Recursive research and synthesis engine.
Keyless GitHub Adapter: Implement adapter for GitHub content access.

License

Apache License 2.0. See LICENSE.

mcp-open-webresearch

Version Issues

Proxy-aware Model Context Protocol (MCP) server for web searching and content extraction.

Designed to be robust and compatible with various network environments, including those using SOCKS and HTTP proxies.

Features

Dynamic Engine Discovery: Engines are loaded dynamically from the src/infrastructure/search/ directory. Adding a new engine requires only a new folder and file, without modifying core logic.
Multi-Engine Search: Aggregates results from Bing, DuckDuckGo, and Brave.
Deep Research (search_deep): Recursive research agent that performs multi-round searching, citation extraction, and answer synthesis.
Ephemeral Downloads: In-memory storage for Deep Search reports using a 100MB bounded LRU cache with 10-minute auto-expiration.
Centralized Throttling: Rate limit management (search and pagination cooldowns) across prioritized engines.
Smart Fetch: Configurable fetching utility (impit) with two operational profiles:
- Browser Mode: Includes modern browser headers (User-Agent, Client Hints) for compatibility with sites requiring browser-standard requests.
- Standard Mode: Uses a minimal HTTP client profile for environments where browser-like identification is not required.
Result Sampling: Optional LLM-based filtering to assess result relevance.
Content Extraction: Webpage visiting and markdown extraction tool (visit_webpage) using a headless browser.
Proxy Support: Full support for SOCKS5, HTTPS, and HTTP proxies.
Configuration: Configurable via environment variables and CLI arguments.
Deployment: Docker images available for production and testing.

Credits

This project includes work from the following contributors:

Manav Kundra: Initial implementation of the server.
Aasee: Added multiple search engines and Docker support.
mzxrai: Core logic for the visit_page tool.

Installation & Quick Start

Docker (Recommended)

Latest Stable Release:

docker pull ghcr.io/rinaldowouterson/mcp-open-webresearch:latest
docker run -p 3000:3000 ghcr.io/rinaldowouterson/mcp-open-webresearch:latest

Test/Debug Image:

docker pull ghcr.io/rinaldowouterson/mcp-open-webresearch:test

Local Installation

To run the server locally (e.g., in Claude Desktop or Cline):

[!NOTE] Replace /absolute/path/to/project with your actual project path.

Configuration (mcp_config.json):

{
  "mcpServers": {
    "open-webresearch": {
      "command": "npm",
      "args": [
        "run",
        "start:sampling",
        "--silent",
        "--prefix",
        "/absolute/path/to/project"
      ],
      "headers": {},
      "disabled": false
    }
  }
}

Remote Server (Streamable HTTP)

Endpoint: http://localhost:3000/mcp

Configuration:

{
  "mcpServers": {
    "open-webresearch": {
      "serverUrl": "http://localhost:3000/mcp",
      "headers": {}
    }
  }
}

Client Configuration & Timeouts

Deep Search processes can take several minutes to complete. Some MCP clients (like Cline and RooCode) have a default timeout of 60 seconds, which will cause the operation to fail.

You MUST configure a higher timeout in your client settings.

Cline (`cline_mcp_settings.json`)

Add the "timeout" parameter (in seconds). Recommended: 1800 (30 minutes).

{
  "mcpServers": {
    "open-webresearch": {
      "disabled": false,
      "timeout": 1800,
      "type": "stdio",
      "command": "npm",
      "args": [
        "run",
        "start:sampling",
        "--silent",
        "--prefix",
        "/absolute/path/to/mcp-open-webresearch"
      ],
      "autoApprove": []
    }
  }
}

RooCode (`mcp_settings.json`)

RooCode also respects the timeout parameter.

{
  "mcpServers": {
    "open-webresearch": {
      "disabled": false,
      "timeout": 1800,
      "command": "npm",
      "args": [
        "run",
        "start:sampling",
        "--silent",
        "--prefix",
        "/absolute/path/to/mcp-open-webresearch"
      ],
      "alwaysAllow": []
    }
  }
}

Antigravity / Windsurf (`mcp_config.json`)

Antigravity / Windsurf handles long-running tools natively, but if they let you configure a timeout, it is best practice to do so.

{
  "mcpServers": {
    "open-webresearch": {
      "command": "npm",
      "args": [
        "run",
        "start:sampling",
        "--silent",
        "--prefix",
        "/absolute/path/to/mcp-open-webresearch"
      ],
      "disabled": false
    }
  }
}

Developer Guide: Adding New Engines

To add a new search engine:

Create Directory: src/infrastructure/search/{engine_name}/
Implement Logic: Create {engine_name}.ts with the fetching/parsing logic.

Export Interface: Create index.ts exporting the SearchEngine interface:

import type { SearchEngine } from "../../../types/search.js";
import { searchMyEngine } from "./my_engine.js";
import { isThrottled } from "../../throttle.js"; // Optional

export const engine: SearchEngine = {
  name: "my_engine",
  search: searchMyEngine,
  isRateLimited: () => isThrottled("my_engine"),
};

Restart: The server will automatically discover and load the new engine.

Build and Run

Locally

# 1. Clone
git clone https://github.com/rinaldowouterson/mcp-open-webresearch.git
cd mcp-open-webresearch

# 2. Install
npm install

# 3. Build & Start
npm run build
npm start

Docker

# Production
docker build -t mcp-websearch .
docker run -p 3000:3000 mcp-websearch

# Testing
npm run test:docker

Testing

Unit & E2E Tests

Uses Vitest for testing. Includes dynamic contract tests for all discovered engines.

npm test

Compliance Tests

Verifies the "Smart Fetch" behavior (User-Agent headers) usage using a local mock server.

npm run test .test/engines/smart_fetch_mode.test.ts

Infrastructure Validation

Validates Docker image builds and basic functionality.

npm run test:infrastructure

Available Scripts

Command	Description
`npm run build`	Compiles TypeScript to `build/` folder.
`npm run watch`	Recompiles on file changes.
`npm run inspector`	Launches MCP inspector UI.
`npm start`	Runs the compiled server.
`npm test`	Runs local tests.
`npm run test:docker`	Runs tests in Docker container.
`npm run test:infrastructure`	Validates docker images.
`npm run generate-certs`	Generates self-signed certificates for testing.

Configuration

Configuration is managed via Environment Variables or CLI arguments.

Variable	Default	Description
`PORT`	`3000`	Server port.
`PUBLIC_URL`	`http://localhost:port`	Public URL for download links.
`ENABLE_CORS`	`false`	Enable CORS.
`CORS_ORIGIN`	`*`	Allowed CORS origin.
`DEFAULT_SEARCH_ENGINES`	`bing,duckduckgo,brave`	Default engines list.
`ENABLE_PROXY`	`false`	Enable proxy support.
`HTTP_PROXY`	-	HTTP Proxy URL.
`HTTPS_PROXY`	-	HTTPS Proxy URL.
`SOCKS5_PROXY`	-	SOCKS5 Proxy URL (Highest Priority).
`SAMPLING`	`false`	Enable result sampling.
`SKIP_IDE_SAMPLING`	`false`	Prefer external API over IDE.
`LLM_BASE_URL`	-	External LLM API base URL.
`LLM_API_KEY`	-	External LLM API key.
`LLM_NAME`	-	External LLM model name.
`LLM_TIMEOUT_MS`	`30000`	Timeout for external LLM calls.
`DEEP_SEARCH_MAX_LOOPS`	`20`	Max research iterations.
`DEEP_SEARCH_RESULTS_PER_ENGINE`	`5`	Results per engine per round.
`DEEP_SEARCH_SATURATION_THRESHOLD`	`0.6`	Threshold to stop research early.
`DEEP_SEARCH_MAX_CITATION_URLS`	`10`	Max URLs to visit for citations.
`DEEP_SEARCH_REPORT_RETENTION_MINUTES`	`10`	Download expiration time (minutes).
`WRITE_DEBUG_TERMINAL`	`false`	Log debug output to stdout.
`WRITE_DEBUG_FILE`	`false`	Log debug output to file.

CLI Arguments

CLI arguments override environment variables.

Argument	Description
`--port <number>`	Port to listen on.
`--debug`	Enable debug logging (stdout).
`--debug-file`	Enable debug logging (file).
`--cors`	Enable CORS.
`--proxy <url>`	Proxy URL (http, https, socks5).
`--engines <items>`	Comma-separated list of engines.
`--sampling`	Enable sampling.
`--no-sampling`	Disable sampling.

Search Pipeline & Scoring

The server uses a multi-stage pipeline to aggregate and refine search results:

1. Multi-Engine Retrieval

Concurrent requests are dispatched to all configured engines (Bing, Brave, DuckDuckGo). Raw results are collected into a single pool.

2. Consensus Scoring & Deduplication

Results are grouped by their canonical URL (protocol/www-agnostic hash).

Deduplication: Multiple entries for the same URL are merged.
Scoring: A consensusScore is calculated for each unique URL:
- Inverted Rank Sum: Sum of inverted ranks ($1/rank$) across engines. Higher placement results in a higher score.
- Engine Boost: Multiplies the sum by the number of unique engines that identified the URL. This prioritizes multi-provider agreement.
Sorting: The final list is sorted by the calculated consensusScore in descending order.

3. LLM Sampling (Optional)

If SAMPLING=true, the top-ranked results are sent to an LLM to evaluate semantic relevance to the query.

Filtering: Sampling acts as a binary filter. It removes results identified as irrelevant (spam, off-topic).
Final Set: The original consensus scores are preserved. Only the composition of the list changes.

LLM Sampling Strategy

When sampling is enabled, the server follows a tiered resolution logic to select which LLM to use:

SKIP_IDE_SAMPLING	IDE Available	API Configured	Resolution
`false` (default)	✅	✅	IDE Sampling
`true`	✅	❌	IDE Sampling
`false`	❌	✅	External API
`true`	✅ OR ❌	✅	External API
`false` OR `true`	❌	❌	No Sampling

[!TIP] You can use a model without API key, the LLM_API_KEY value is optional.

[!IMPORTANT] Deep Search Compatibility: The search_deep tool strictly requires LLM capability (either via IDE or API). If neither is available, the tool will appear in the MCP list but will throw an error upon execution.

Tools Documentation

`search_deep`

Recursive research agent for deep investigation. Searches multiple sources, extracts citations, and synthesizes a comprehensive answer.

Requires LLM Sampling capability.

Input:

{
  "objective": "Deep research goal",
  "max_loops": 3,
  "results_per_engine": 5,
  "max_citation_urls": 10,
  "engines": ["bing", "brave"],
  "attach_context": false
}

Output: A structured Markdown report including a reference list. If configured, a Download URL at the top of the output permits downloading the results as a file.

`search_web`

Performs a search across configured engines.

Input:

{
  "query": "search query",
  "max_results": 10,
  "engines": ["bing", "brave"],
  "sampling": true
}

`visit_webpage`

Visits a URL and returns markdown content.

Input:

{
  "url": "https://example.com/article",
  "capture_screenshot": false
}

`set_engines`

Updates default search engines.

Input:

{
  "engines": ["duckduckgo", "brave"]
}

`get_engines`

Returns configured search engines.

`set_sampling`

Enables or disables result sampling.

Input:

{
  "enabled": true
}

`get_sampling`

Returns current sampling status.

📥 Ephemeral Downloads

Deep Search results are served via an in-memory buffer cache.

Storage: Reports are stored as Buffer objects in the C++ heap to avoid V8 string memory limits.
Expiration: Each individual entry expires exactly 10 minutes after creation. Access operations (get) do not extend the time-to-live (TTL).
Memory Safety: The cache is bounded by a 100MB ceiling. When the limit is reached, a Least Recently Used (LRU) eviction policy removes the oldest entries.
URL Configuration: Link generation depends on the PUBLIC_URL variable to ensure accessible download endpoints in proxied environments.

Roadmap

Deep Search: Recursive research and synthesis engine.
Keyless GitHub Adapter: Implement adapter for GitHub content access.

License

Apache License 2.0. See LICENSE.

mcp-open-webresearch

mcp-open-webresearch

Features

Credits

Installation & Quick Start

Docker (Recommended)

Local Installation

Remote Server (Streamable HTTP)

Client Configuration & Timeouts

Cline (cline_mcp_settings.json)

RooCode (mcp_settings.json)

Antigravity / Windsurf (mcp_config.json)

Developer Guide: Adding New Engines

Build and Run

Locally

Docker

Testing

Unit & E2E Tests

Compliance Tests

Infrastructure Validation

Available Scripts

Configuration

CLI Arguments

Search Pipeline & Scoring

1. Multi-Engine Retrieval

2. Consensus Scoring & Deduplication

3. LLM Sampling (Optional)

LLM Sampling Strategy

Tools Documentation

search_deep

search_web

visit_webpage

set_engines

get_engines

set_sampling

get_sampling

📥 Ephemeral Downloads

Roadmap

License

mcp-open-webresearch

Features

Credits

Installation & Quick Start

Docker (Recommended)

Local Installation

Remote Server (Streamable HTTP)

Client Configuration & Timeouts

Cline (cline_mcp_settings.json)

RooCode (mcp_settings.json)

Antigravity / Windsurf (mcp_config.json)

Developer Guide: Adding New Engines

Build and Run

Locally

Docker

Testing

Unit & E2E Tests

Compliance Tests

Infrastructure Validation

Available Scripts

Configuration

CLI Arguments

Search Pipeline & Scoring

1. Multi-Engine Retrieval

2. Consensus Scoring & Deduplication

3. LLM Sampling (Optional)

LLM Sampling Strategy

Tools Documentation

search_deep

search_web

visit_webpage

set_engines

get_engines

set_sampling

get_sampling

📥 Ephemeral Downloads

Roadmap

License

Cline (`cline_mcp_settings.json`)

RooCode (`mcp_settings.json`)

Antigravity / Windsurf (`mcp_config.json`)

`search_deep`

`search_web`

`visit_webpage`

`set_engines`

`get_engines`

`set_sampling`

`get_sampling`

Cline (`cline_mcp_settings.json`)

RooCode (`mcp_settings.json`)

Antigravity / Windsurf (`mcp_config.json`)

`search_deep`

`search_web`

`visit_webpage`

`set_engines`

`get_engines`

`set_sampling`

`get_sampling`