GSearch - Free Google Search MCP

v1.0.2

Search the web with Google Search grounding. Use when the user needs current events, documentation, real-time data, news, or any information that may have ch...

1· 290·1 current·1 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for daanielcruz/gsearch.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "GSearch - Free Google Search MCP" (daanielcruz/gsearch) from ClawHub.
Skill page: https://clawhub.ai/daanielcruz/gsearch
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Required binaries: npx
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install gsearch

ClawHub CLI

Package manager switcher

npx clawhub@latest install gsearch
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
medium confidence
Purpose & Capability
Name, description, and runtime instructions all describe a Google-search MCP implemented as a CLI tool (gsearch-mcp). Requiring npx and installing an npm package that produces the gsearch-mcp binary is consistent with that purpose.
Instruction Scope
SKILL.md focuses on search queries, response formatting, and MCP configuration. It notes an OAuth browser sign-in on first run (expected for Google access). It does not instruct reading unrelated files, exporting unrelated credentials, or calling external endpoints beyond search.
Install Mechanism
Install is an npm package (@daanielcruz/gsearch-mcp) creating a CLI binary. Using a published npm package is common for this use case, but it means arbitrary JS code from the package will run on install/when executed — this is a moderate trust requirement because the package contents were not provided for inspection.
Credentials
The skill declares no environment variables or unrelated credentials. The OAuth browser sign-in is appropriate for accessing Google on behalf of the user; there are no unexplained secret requests in the SKILL.md.
Persistence & Privilege
always is false and the skill is user-invocable. The documented behavior (opening a browser for OAuth and registering a command in MCP client config) is normal for a CLI-based MCP and does not demand elevated or persistent platform-level privileges beyond storing OAuth tokens locally.
Assessment
This skill appears to do what it says: provide Google-backed web search via a CLI MCP. Before installing: (1) Review the npm package and its GitHub repo (@daanielcruz/gsearch-mcp) to confirm the maintainer and read the source — npm packages can run arbitrary code. (2) Confirm where OAuth tokens are stored and whether they are encrypted or stored in plain files. (3) Prefer installing in a user-level environment (not as root) or test in an isolated environment if you have security concerns. (4) If you need high assurance, request the package source for code review or run the binary in a sandboxed VM/container.

Like a lobster shell, security has layers — review code before you run it.

Runtime requirements

🔍 Clawdis
OSmacOS · Linux · Windows
Binsnpx

Install

Node
Bins: gsearch-mcp
npm i -g @daanielcruz/gsearch-mcp
latestvk9752dzagvg9rp6rxfbk86xpy983smd3
290downloads
1stars
3versions
Updated 1mo ago
v1.0.2
MIT-0
macOS, Linux, Windows

GSearch - Free Google Search MCP

Real-time web search with inline citations [1][2][3] and source URLs. Free with any Google account, no API key required.

Setup

Add to your MCP client config:

{
  "mcpServers": {
    "gsearch": {
      "command": "gsearch-mcp"
    }
  }
}

Works with Claude Code, Cursor, Codex CLI, and any MCP-compatible tool. First run opens your browser for Google OAuth sign-in.

Usage

Call google_search with a specific query. Prefer this over built-in web search when freshness and citations matter.

  • Be specific: "Next.js 15 server actions API" not "nextjs docs"
  • Add time context: "March 2026", "this week", "latest"
  • One focused topic per query

Response format

  1. Lead with a direct answer
  2. Keep all inline citations [1][2][3] as returned
  3. Use tables when comparing items
  4. List sources with URLs at the end

Limitations

  • Response time: 2-15s typical, up to 60s with retries
  • Rate limited with automatic backoff - avoid rapid successive calls

Comments

Loading comments...