← Back to directory
K

Kindly Web Search MCP Server

Community
Web search plus robust content retrieval for AI coding tools and agents
GitHub source repository ↗
★ 385 Stars Category · Dev Tools Very popular
59FMRS · C

Kindly Web Search MCP Server is a content-quality-focused search MCP service: specialized API integrations deliver full StackExchange, GitHub Issues/Discussions, Wikipedia and arXiv content to the model in one call, while a headless Chromium browser parses other pages in real time, with multiple search providers and proxy support. It suits users who want less scraping round-tripping and better AI coding context, but it requires a local browser and at least one search key, and its remote HTTP mode is unauthenticated and unencrypted by default, so deployment needs extra protection.

Reliability
9/20
Security and permissions
13/20
Maintenance
11/20
Documentation
14/20
Setup experience
12/20
Read the FMRS scoring method →

Kindly Web Search MCP Server is part of the Shelpuk AI Technology Consulting agentic suite and provides web search plus robust content retrieval for AI coding tools (Claude Code, Codex, Cursor, GitHub Copilot, Gemini, etc.) and AI agents (Claude Desktop, OpenClaw, Hermes, etc.). Rather than returning only titles, links and snippets, it retrieves the full conversation — questions, answers, comments, reactions and metadata — and integrates directly with the StackExchange, GitHub Issues, GitHub Discussions, Wikipedia and arXiv APIs, while any other URL is parsed in real time with a local headless Chromium browser. Search supports Serper, SerpBase, Tavily, SearXNG, Sofya and You.com with priority-based fallback. It targets AI coding and agent workflows that want usable context in a single call instead of an extra scraping round trip.

Tools

web_search
Takes a query and num_results (default 3) and returns title, link, snippet and best-effort Markdown page_content, with per-result diagnostics when diagnostics are enabled.
get_content
Takes a URL and returns best-effort Markdown page_content, with top-level diagnostics when diagnostics are enabled.

Setup

1) Install uvx (curl script on macOS/Linux, PowerShell script on Windows). 2) Install a Chromium-based browser (Chrome/Chromium/Edge/Brave) for page_content extraction; specialized sources still work without it. 3) Set at least one search provider key: SERPER_API_KEY, SERPBASE_API_KEY, TAVILY_API_KEY, SEARXNG_BASE_URL, SOFYA_API_KEY or YDC_API_KEY. 4) A read-only GITHUB_TOKEN is highly recommended for better GitHub Issue extraction. 5) Configure your MCP client with the run command: uvx --from git+https://github.com/Shelpuk-AI-Technology-Consulting/kindly-web-search-mcp-server kindly-web-search-mcp-server start-mcp-server. The first run may take 30-60 seconds to build the environment; prewarm it in a terminal if your client times out.

claude_desktop_config.json
{
  "mcpServers": {
    "kindly-web-search": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/Shelpuk-AI-Technology-Consulting/kindly-web-search-mcp-server",
        "kindly-web-search-mcp-server",
        "start-mcp-server"
      ],
      "env": {
        "SERPER_API_KEY": "${SERPER_API_KEY}",
        "TAVILY_API_KEY": "${TAVILY_API_KEY}",
        "SEARXNG_BASE_URL": "${SEARXNG_BASE_URL}",
        "GITHUB_TOKEN": "${GITHUB_TOKEN}",
        "KINDLY_BROWSER_EXECUTABLE_PATH": "${KINDLY_BROWSER_EXECUTABLE_PATH}"
      }
    }
  }
}

Fit and risk

Best for

  • Users of AI coding tools who need full discussion/documentation context rather than just URLs
  • Clients that support MCP, such as Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI and Claude Desktop
  • Teams deploying a remote MCP service via Docker and Streamable HTTP
  • Environments that must egress through a proxy (HTTP/SOCKS5)

Not for

  • Users who cannot install a Chromium-based browser on the MCP client machine and mainly need arbitrary-site page_content extraction
  • Users who want to expose the remote HTTP port directly to the public internet (it is unauthenticated and unencrypted by default)
  • Users unwilling to provide any search provider API key

Required permissions

  • Reads a search provider API key from the environment (one of Serper/SerpBase/Tavily/SearXNG/Sofya/You.com)
  • Optional GITHUB_TOKEN (can be read-only and limited to public repositories) for GitHub Issues/Discussions extraction
  • Launches a local headless Chromium browser process to fetch arbitrary web pages
  • Optional outbound network access via proxy env vars (HTTP_PROXY/HTTPS_PROXY/ALL_PROXY) or KINDLY_CHROME_PROXY

Risks and side effects

  • HTTP transports are unauthenticated and unencrypted by default; exposing the port publicly is risky
  • get_content fetches any URL it is given and runs from the server's network location, which can reach internal resources
  • Browser reuse shares cookies, local storage, cache and user-agent state across requests
  • API keys stored in client config files may leak
  • Failed or timed-out fetches may return partial content or an error message

Troubleshooting

  1. No browser executable found: install Chrome/Chromium/Edge and set KINDLY_BROWSER_EXECUTABLE_PATH
  2. Failed to connect to browser: raise KINDLY_NODRIVER_RETRY_ATTEMPTS, KINDLY_NODRIVER_DEVTOOLS_READY_TIMEOUT_SECONDS and KINDLY_HTML_TOTAL_TIMEOUT_SECONDS, and set NO_PROXY=localhost,127.0.0.1
  3. page_content shows TimeoutError: increase KINDLY_TOOL_TOTAL_TIMEOUT_SECONDS and KINDLY_TOOL_TOTAL_TIMEOUT_MAX_SECONDS, optionally set KINDLY_WEB_SEARCH_MAX_CONCURRENCY=1
  4. HTTP deployment returns 421/403: widen the allowlists with FASTMCP_ALLOWED_HOSTS and FASTMCP_ALLOWED_ORIGINS
  5. web_search fails with no provider key: set at least one search provider environment variable

Use cases

An AI coding assistant retrieving the full StackOverflow thread and answers when debugging a complex error
Fetching GitHub Issue/Discussion bodies, comments and reactions in a single agent call
Pulling arXiv paper content and Wikipedia entries for research-oriented tasks
Supplementing coding tasks with up-to-date API and package documentation
Running searches through a self-hosted SearXNG instance in restricted environments

Supported clients

Claude CodeFull support
CodexFull support
Gemini CLIFull support
Claude DesktopFull support
CursorFull support
OpenClawFull support
AntigravityFull support
GitHub Copilot (VS Code)Full support