Add a Parallel Search MCP walkthrough

README.md

@@ -729,6 +729,7 @@ Full setup walkthroughs in [`mcp-servers/`](./mcp-servers/). New to MCP? [`mcp-s
 | **Sequential Thinking** | Step-by-step reasoning that breaks complex problems into manageable steps | [sequential-thinking.md](./mcp-servers/sequential-thinking.md) |
 | **Memory** | Persistent context across sessions | [memory.md](./mcp-servers/memory.md) |
 | **Playwright** | Browser automation — interaction, scraping, testing, accessibility | [playwright.md](./mcp-servers/playwright.md) |
+| **Parallel Search** | Web search and page retrieval through a hosted remote server | [parallel-search.md](./mcp-servers/parallel-search.md) |
 
 See [`mcp-servers/README.md`](./mcp-servers/README.md) for the comparison matrix, install commands, and troubleshooting.
 

mcp-servers/README.md

@@ -2,7 +2,7 @@
 
 ## Overview
 
-This directory contains comprehensive documentation for the four core Model Context Protocol (MCP) servers that enhance Claude Code's capabilities.
+This directory contains comprehensive documentation for five Model Context Protocol (MCP) servers that enhance Claude Code's capabilities.
 
 ---
 
@@ -14,6 +14,7 @@ This directory contains comprehensive documentation for the four core Model Cont
 | **Sequential Thinking** | Advanced reasoning | • Problem decomposition<br>• Multi-step planning<br>• Structured analysis<br>• Decision support | [sequential-thinking.md](./sequential-thinking.md) |
 | **Memory** | Persistent context | • Cross-session memory<br>• Project preferences<br>• Historical context<br>• Knowledge retention | [memory.md](./memory.md) |
 | **Playwright** | Browser automation | • Web scraping<br>• Automated testing<br>• Screenshot capture<br>• Device emulation | [playwright.md](./playwright.md) |
+| **Parallel Search** | Web research | • Web search<br>• Page retrieval<br>• Remote HTTP connection | [parallel-search.md](./parallel-search.md) |
 
 > **Tip:** Keep only 3–6 MCP servers active at a time — every connected server adds tools to Claude's context.
 
@@ -36,6 +37,7 @@ This directory contains comprehensive documentation for the four core Model Cont
 - **Serena**: Analyze code structure and dependencies
 - **Sequential Thinking**: Break down complex problems
 - **Playwright**: Gather web data
+- **Parallel Search**: Find sources and retrieve page content
 
 ---
 
@@ -64,11 +66,12 @@ Use case: Automated testing workflows
 - Memory remembers test patterns and credentials
 ```
 
-**4. All Four Together**
+**4. All Five Together**
 ```
 Use case: Full-stack development
 - Serena for backend code intelligence
 - Playwright for frontend testing
+- Parallel Search for web research
 - Sequential Thinking for planning
 - Memory for project knowledge
 ```
@@ -220,6 +223,7 @@ playwright: ✓ Connected
 - **[Sequential Thinking](./sequential-thinking.md)** - For complex problem-solving
 - **[Memory](./memory.md)** - For persistent project knowledge
 - **[Playwright](./playwright.md)** - For web automation needs
+- **[Parallel Search](./parallel-search.md)** - For web search and page retrieval
 
 ---
 

mcp-servers/parallel-search.md

@@ -0,0 +1,68 @@
+# Parallel Search MCP Server
+
+*~3 min read*
+
+## Overview
+
+> **Mental model:** Think of Parallel Search MCP as a web research adapter for Claude Code: one remote connection provides both search results and readable page content.
+
+The hosted server exposes two tools:
+
+| Tool | Purpose |
+|---|---|
+| `web_search` | Search the web for relevant sources |
+| `web_fetch` | Retrieve content from a specific URL |
+
+## Prerequisites
+
+- Claude Code CLI
+- Internet access
+
+The default endpoint works without an API key.
+
+## Installation
+
+### Global Installation
+
+Add the server for all projects:
+
+```bash
+claude mcp add --transport http -s user parallel-search https://search.parallel.ai/mcp
+```
+
+### Local Installation
+
+Add it only to the current project:
+
+```bash
+claude mcp add --transport http -s local parallel-search https://search.parallel.ai/mcp
+```
+
+## Usage
+
+Ask Claude to search and open useful results:
+
+```text
+Search the web for the current Claude Code MCP documentation, open the most relevant official source, and summarize the setup steps with links.
+```
+
+## Verification
+
+Confirm that the server is connected:
+
+```bash
+claude mcp list
+```
+
+The output should list `parallel-search` as connected. You can also type `/mcp`
+inside Claude Code to inspect the available tools.
+
+## Troubleshooting
+
+If the server does not connect:
+
+1. Confirm the URL is exactly `https://search.parallel.ai/mcp`.
+2. Remove and add the entry again: `claude mcp remove parallel-search`.
+3. Check that your network permits outbound HTTPS connections.
+
+See the [official Parallel Search MCP documentation](https://docs.parallel.ai/integrations/mcp/search-mcp) for access options and client-specific setup.