Agent Tools
Provides 9 utility APIs for AI agents covering DNS, email, IP, QR, WHOIS, URL health, headers, text extraction, and tech detection, with x402 micropayments on Base.
README
Agent Tools
Agent-native utility API bundle. 9 endpoints exposed as both REST API and MCP server (streamable-http), with x402 micropayments. Live at tools.beethoven2024.com.
For AI agents (Claude, Cursor, custom): Add this MCP server to your config and use 9 utility tools instantly.
For developers: Standard REST API. Use any subdomain example below as a working sample.
Try it now
As MCP server (Claude, Cursor, agents)
{
"mcpServers": {
"agent-tools": {
"url": "https://tools.beethoven2024.com/mcp"
}
}
}
As REST API
curl https://tools.beethoven2024.com/health
# {"status":"ok","version":"0.1.0"}
curl "https://tools.beethoven2024.com/v1/dns/health?domain=example.com"
# Full DNS + mail security audit (requires x402 payment)
What's in the bundle
| Tool | Endpoint | What it does | Price |
|---|---|---|---|
qr_generate |
/v1/qr/generate |
QR code generation (PNG/SVG, 64–2048px) | $0.001 |
dns_health |
/v1/dns/health |
DNS records + SPF/DMARC mail security audit | $0.003 |
email_validate |
/v1/email/validate |
Format, MX, disposable domain, role account, SMTP probe | $0.005 |
ip_lookup |
/v1/ip/lookup |
Geolocation, ISP, proxy/hosting detection, reverse DNS | $0.005 |
url_health |
/v1/url/health |
Status code, redirect chain, SSL cert, response time | $0.003 |
whois_lookup |
/v1/whois/lookup |
Registrar, creation/expiration dates, nameservers, status | $0.005 |
headers_analyze |
/v1/headers/analyze |
HTTP security headers audit with 0–100 score (HSTS, CSP, etc.) | $0.003 |
extract_text |
/v1/extract/text |
Clean text extraction from webpages (strips HTML, nav, ads) | $0.005 |
tech_detect |
/v1/tech/detect |
Website technology stack — frameworks, CMS, CDN, analytics | $0.005 |
Free: /health, /, /docs (OpenAPI), and MCP protocol methods initialize / tools/list / ping. All 9 tools require x402 payment via both MCP (tools/call) and REST (/v1/*).
Why these endpoints
Every endpoint solves a problem agents can't easily do locally:
- Network infrastructure (DNS, WHOIS, headers, URL health, IP) — needs internet probes, can't be done in-process
- Specialized parsing (QR, email validation with SMTP, tech detection) — domain-specific logic agents shouldn't reimplement
- Web extraction (text extraction) — clean text from messy HTML is hard
8 of 9 endpoints are fully self-hosted (zero upstream API cost). Only IP geolocation uses an external free-tier API.
Payment
x402 (machine-to-machine micropayments)
Every paid endpoint returns HTTP 402 Payment Required with the price spec when no payment is included. x402-capable agents handle this automatically:
- Agent calls endpoint → receives 402 with payment_requirements (network, amount, recipient address)
- Agent signs EIP-3009
transferWithAuthorizationfor USDC on Base - Agent retries with
X-PAYMENTheader - Service verifies payment via facilitator, executes, returns result
Settlement happens on-chain on Base. Sub-2-second confirmation.
Runs on Base mainnet by default. Set AGENT_TOOLS_TESTNET=true to use Base Sepolia with the public testnet facilitator.
Traditional billing
Coming soon: API key + Stripe usage billing for human developers and businesses.
Tech stack
- Python 3.12 + FastAPI
- MCP SDK (
mcpv1.27) — streamable-http transport - x402 SDK (
x402[fastapi,evm]>=v2.11) — USDC micropayments on Base - 8/9 endpoints self-hosted (no upstream API cost)
- Single-process deploy, ~62MB memory footprint
Self-host
Fly.io
git clone https://github.com/WingedGuardian/agent-tools.git
cd agent-tools
fly launch --copy-config
fly secrets set AGENT_TOOLS_PAY_TO=0xYourAddress AGENT_TOOLS_TESTNET=false
fly deploy
Docker
git clone https://github.com/WingedGuardian/agent-tools.git
cd agent-tools
cp .env.example .env
# Edit .env with your wallet address
docker compose up -d
Manual
pip install ".[all]"
AGENT_TOOLS_PAY_TO=0xYourAddress uvicorn agent_tools.app:app --host 0.0.0.0 --port 8000
Configuration
| Variable | Required | Default | Description |
|---|---|---|---|
AGENT_TOOLS_PAY_TO |
No | (empty) | EVM wallet address for receiving USDC payments. Leave empty to run free/dev mode. |
AGENT_TOOLS_TESTNET |
No | true |
true = Base Sepolia testnet (default — works with public x402.org facilitator). false = Base mainnet (requires mainnet-capable facilitator). |
Discovery / registries
- GitHub: https://github.com/WingedGuardian/agent-tools
- Smithery: (pending registration)
- Glama: (pending registration)
- x402-list: (pending registration)
- npm: (pending publish)
Keywords
x402, micropayments, USDC, Base, agent-native, MCP, Model Context Protocol, streamable-http, agentic, AI agent tools, DNS lookup API, WHOIS API, email validation API, QR code API, URL health API, web scraping API, tech detection API, security headers analyzer
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。