MCTS安全扫描器 是 AI Skill Hub 本期精选MCP工具之一。综合评分 7.5 分,整体质量较高。我们推荐使用将其纳入你的 AI 工具库,帮助提升工作效率。
MCTS安全扫描器 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
MCTS安全扫描器 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/MCP-Audit/MCTS
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"mcts-----": {
"command": "npx",
"args": ["-y", "mcts"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 MCTS安全扫描器 执行以下任务... Claude: [自动调用 MCTS安全扫描器 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"mcts_____": {
"command": "npx",
"args": ["-y", "mcts"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
Model Context Threat Scanner
Security scanner for Model Context Protocol (MCP) servers — the programs that give AI assistants access to tools, files, databases, and APIs.
Run one command to find permission issues, injection risks, attack chains, and more. Works locally, in CI, with no cloud account required.
New to MCP or MCTS? See the documentation index and glossary.
mcts scan ./server.py # single entrypoint
mcts scan ./ # entire repository
MCTS is alpha software with a local-first MCP security pipeline — no cloud account required for standard scans. Full reference: Security checks · CLI.
Recommended — isolated tool install (does not touch your app venv):
uvx mcp-mcts scan ./server.py
pipx install mcp-mcts
uv tool install mcp-mcts
Distribution name ismcp-mcts(the genericmctsname is already taken on PyPI). The import package and CLI remainmcts.
From PyPI in a dedicated environment (not your application .venv):
pip install mcp-mcts
pip install "mcp-mcts[mcp]" # live probing + fuzzing
pip install "mcp-mcts[api]" # REST API (`mcts serve`)
pip install "mcp-mcts[llm]" # LLM-as-judge / --llm-triage (install separately; not in [all])
pip install "mcp-mcts[semgrep]" # Semgrep SAST adapter (--semgrep; also needs semgrep CLI)
Avoid pip install mcp-mcts[all] inside your app's dev venv — it can conflict with pinned dependencies.
| Install | Use case |
|---|---|
uvx mcp-mcts | One-off scan, no install |
pipx install mcp-mcts | Global isolated CLI |
pip install mcp-mcts[mcp] | Live probing in a dedicated venv |
pip install mcp-mcts[all] | All extras except LLM (no litellm) |
pip install 'mcp-mcts[all,llm]' | Everything including --llm-judge / --llm-triage |
From source (contributors):
git clone https://github.com/MCP-Audit/MCTS.git
cd MCTS
uv sync --all-extras
Scan the included vulnerable MCP server:
uv run mcts scan examples/vulnerable-mcp-server/server.py

<details> <summary>Example terminal output</summary>
$ mcts scan examples/vulnerable-mcp-server/server.py
[✓] Discovering tools...
[✓] Mapping permissions...
[✓] Detecting attack chains...
[✓] Generating report...
==================== MCTS Security Report ====================
Overall Score: 5/100 (CRITICAL)
Risk Index: 100/100
Scoring basis: 3 Critical, 7 High, 2 Medium (12 scorable findings)
Severity Summary Top Findings
● Critical 4 [1] CRITICAL Destructive tool: delete_all_users
● High 7 [2] CRITICAL Read → exfiltration attack chain possible
● Medium 2 ...
</details>
| Capability | How |
|---|---|
| REST API | mcts serve — 10 scan endpoints (--extra api) |
| MCP server mode | mcts-mcp — scan_mcp_target, explain_finding, compare_baselines for IDE agents |
| Readiness (non-security) | mcts readiness — HEUR-001–020 + optional OPA/LLM |
| Protocol fuzzing | mcts fuzz — safe read-only probes by default |
| Surface subcommands | mcts scan-prompts, scan-resources, scan-instructions (surface-scoped analyzers; no supply-chain noise) |
| Python API | from mcts import Scanner, ScanConfig |
MCTS是一个有用的MCP安全扫描工具,能够有效检测模型安全威胁
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ Apache 2.0 — 宽松开源协议,可商用,需保留版权声明和 NOTICE 文件,含专利授权条款。
经综合评估,MCTS安全扫描器 在MCP工具赛道中表现稳健,质量良好。如果你已有明确的使用需求,可以直接上手体验;如果还在评估阶段,建议对比同类工具后再做决策。
| 原始名称 | MCTS |
| 原始描述 | 开源MCP工具:MCTS (Model Context Threat Scanner) is a local-first security scanner for MCP se。⭐17 · Python |
| Topics | ai-agentsai-safetymcpsecurity |
| GitHub | https://github.com/MCP-Audit/MCTS |
| License | Apache-2.0 |
| 语言 | Python |
收录时间:2026-06-09 · 更新时间:2026-06-09 · License:Apache-2.0 · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端