by google
⚠️ Third-Party Software Notice
This skill is third-party open-source software developed and hosted independently on GitHub. SkillsLLM is an informational directory and does not control or maintain the underlying repository.
Any security checks, ratings, or warnings displayed by SkillsLLM are automated and limited in scope. They do not constitute a security certification or guarantee that the software is safe, error-free, or free from malicious code, vulnerabilities, compromised dependencies, or prompt-injection risks.
Review the source code, permissions, dependencies, and configuration before installing or running any third-party skill. Use is at your own risk. To the maximum extent permitted by applicable law, SkillsLLM is not liable for losses arising from third-party software.
# Add to your Claude Code skills
git clone https://github.com/google/mcp-securityGuides for using ai agents skills like mcp-security.
Last scanned: 9/1/2026
{
"issues": [
{
"file": "extensions/google-secops/skills/setup-gemini-cli/SKILL.md",
"line": 18,
"type": "remote-install",
"message": "Install command (remote install script piped to a shell — review the source before running): \"curl -LsSf https://astral.sh/uv/install.sh | sh\"",
"severity": "medium"
}
],
"status": "PASSED",
"scannedAt": "2026-09-01T09:11:45.813Z",
"npmAuditRan": true,
"pipAuditRan": true,
"promptInjectionRan": true
}This repository contains Model Context Protocol (MCP) servers that enable MCP clients (like Claude Desktop or the cline.bot VS Code extension) to access Google's security products and services:
For the new Remote MCP Server, please see the launch announcement and the setup guide.
Each server can be enabled and run separately, allowing flexibility for environments that don't require all capabilities.
Comprehensive documentation is available in the docs folder. You can:
The documentation covers:
To get started with the documentation, see docs/index.md.
The server uses Google's authentication. Make sure you have either:
gcloud auth application-default loginEach MCP server can be installed and used as a standalone package.
You can install the packages using uv tool install (recommended):
# Install packages
uv tool install google-secops-mcp
uv tool install gti-mcp
uv tool install scc-mcp
uv tool install secops-soar-mcp
Alternatively, you can use pip:
pip install google-secops-mcp
pip install gti-mcp
pip install scc-mcp
pip install secops-soar-mcp
After installation, you can run the servers directly using uvx:
# Run SecOps MCP server
uvx --from google-secops-mcp secops_mcp
# Run GTI MCP server
uvx gti_mcp
# Run SCC MCP server
uvx scc_mcp
# Run SecOps SOAR MCP server (with optional integrations)
uvx secops_soar_mcp --integrations CSV,OKTA
With environment variables:
CHRONICLE_PROJECT_ID="your-project-id" \
CHRONICLE_CUSTOMER_ID="01234567-abcd-4321-1234-0123456789ab" \
CHRONICLE_REGION="us" \
uvx secops_mcp
You can configure MCP clients to use the installed packages with uvx. Here's an example configuration:
{
"mcpServers": {
"secops": {
"command": "uvx",
"args": [
"--from",
"google-secops-mcp",
"secops_mcp"
],
"env": {
"CHRONICLE_PROJECT_ID": "your-project-id",
"CHRONICLE_CUSTOMER_ID": "01234567-abcd-4321-1234-0123456789ab",
"CHRONICLE_REGION": "us"
}
},
"gti": {
"command": "uvx",
"args": [
"gti_mcp"
],
"env": {
"VT_APIKEY": "0123456789abcdef0123456789abcdef0123456789abcdef0123456789abcdef"
}
},
"scc-mcp": {
"command": "uvx",
"args": [
"scc_mcp"
],
"env": {}
},
"secops-soar": {
"command": "uvx",
"args": [
"secops_soar_mcp",
"--integrations",
"CSV,OKTA"
],
"env": {
"SOAR_URL": "https://yours-here.siemplify-soar.com:443",
"SOAR_APP_KEY": "01234567-abcd-4321-1234-0123456789ab"
}
}
}
}
You can also use environment files with uvx:
{
"mcpServers": {
"secops": {
"command": "uvx",
"args": [
"--env-file",
"/path/to/.env",
"secops_mcp"
]
}
}
}
The MCP servers from this repo can be used with the following clients
The configuration for Claude Desktop and Cline is the same (provided below for uv and pip). We use the stdio transport.
Please refer to the README file for both - locally running the prebuilt agent and Cloud Run deployment.
MCP clients all use the same JSON configuration format (see the MCP Server Configuration Reference), but they expect the file in different locations.
| Client Application | Scope | macOS / Linux Location | Windows Location | Notes |
|---|---|---|---|---|
| Gemini CLI | Global | ~/.gemini/settings.json |
%USERPROFILE%\.gemini\settings.json |
File must include mcpServers. Confirmed in Google Security Ops post. |
| Claude Desktop | Global | ~/Claude/claude_desktop_config.json |
%USERPROFILE%\Claude\claude_desktop_config.json |
Config accessible via Claude > Settings > Developer > Edit Config. |
| Claude Code | Global | ~/.claude.json |
%USERPROFILE%\.claude.json |
Primary config file for Claude Code CLI and extensions. |
| Cursor IDE (Global) | Global | ~/.cursor/mcp.json |
%USERPROFILE%\.cursor\mcp.json |
Enables MCP servers globally across all projects. |
| Cursor IDE (Project) | Project | <project-root>/.cursor/mcp.json |
<project-root>/.cursor/mcp.json |
Workspace/project-specific config file. |
| VS Code (Workspace) | Workspace | <project-root>/.vscode/mcp.json |
<project-root>/.vscode/mcp.json |
Workspace-level config used when an MCP extension (like Cline) is installed. Overrides global config if present. |
| Cline (VS Code Ext.) | Global | Inside VS Code extension data | %APPDATA%\Code\User\globalStorage\<extension-id>\settings\cline_mcp_settings.json |
Exact path varies by VS Code variant and platform. <extension-id> corresponds to the installed extension folder (e.g., saoudrizwan.claude-dev). |
%USERPROFILE% → C:\Users\<username>%APPDATA% → C:\Users\<username>\AppData\Roaming<project-root> → folder opened in VS Code or IDE for the project<extension-id> → name of the installed extension folder (e.g., saoudrizwan.claude-dev for Claude/Cline)If you use multiple MCP clients, you can maintain a single config file and symlink it into each expected location. This avoids drift and keeps your server definitions consistent.
{
"mcpServers": {
"secops": {
"command": "uv",
"args": [
"--directory",
"/path/to/the/repo/server/secops/secops_mcp",
"run",
"server.py"
],
"env": {
"CHRONICLE_PROJECT_ID": "your-project-id",
"CHRONICLE_CUSTOMER_ID": "01234567-abcd-4321-1234-0123456789ab",
"CHRONICLE_REGION": "us"
}
},
"secops-soar": {
"command": "uv",
"args": [
"--directory",
"/path/to/the/repo/server/secops-soar/secops_soar_mcp",
"run",
"server.py",
"--integrations",
"CSV,OKTA"
],
"env": {
"SOAR_URL": "https://yours-here.siemplify-soar.com:443",
"SOAR_
mcp-security is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by google. It has 521 GitHub stars.
Yes. mcp-security passed SkillsLLM's automated security scan — a dependency vulnerability audit plus prompt-injection heuristics — with no high-severity issues. You can read the full report in the Security Report section on this page.
Clone the repository with "git clone https://github.com/google/mcp-security" and add it to your Claude Code skills directory (see the Installation section above).
mcp-security is primarily written in Python. It is open-source under google on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh mcp-security against similar tools.
No comments yet. Be the first to share your thoughts!