13 production-ready tools for AI agents — one API key, zero complexity.
Docs · Playground · OpenAPI Spec · npm
Building AI agents that interact with the real world? You'd normally need:
| Without Agent Toolbox | With Agent Toolbox |
|---|---|
| 6+ API keys (Google, OpenWeatherMap, Yahoo Finance, ...) | 1 API key |
| Different auth methods per service | Unified Bearer auth |
| Parse different response formats | Consistent JSON responses |
| Handle rate limits per provider | Built-in rate limiting |
| No caching, redundant calls | Smart LRU caching |
| Pay $50-200+/mo for APIs | Free tier: 1,000 calls/mo |
curl -X POST https://api.toolboxlite.com/v1/auth/register \
-H "Content-Type: application/json" \
-d '{"email": "you@example.com"}'curl -X POST https://api.toolboxlite.com/v1/search \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"query": "best AI agent frameworks"}'No credit card. No setup. 13 endpoints ready to use.
| Endpoint | Description | Cache |
|---|---|---|
POST /v1/search |
Web search via DuckDuckGo | 5 min |
POST /v1/extract |
Extract content from any URL | 5 min |
POST /v1/screenshot |
Capture webpage screenshots | 1 hr |
POST /v1/weather |
Weather & forecast | 15 min |
POST /v1/finance |
Stock quotes & exchange rates | 15 min |
POST /v1/validate-email |
Email validation (MX + SMTP) | 1 hr |
POST /v1/translate |
Translate 100+ languages | 1 hr |
POST /v1/geoip |
IP geolocation lookup | 1 hr |
POST /v1/news |
News article search | 5 min |
POST /v1/whois |
Domain WHOIS lookup | 1 hr |
POST /v1/dns |
DNS record queries | 5 min |
POST /v1/pdf-extract |
Extract text from PDFs | 1 hr |
POST /v1/qr |
Generate QR codes | 24 hr |
Use Agent Toolbox as an MCP server in Claude Desktop, Cursor, or Windsurf:
npm install -g agent-toolbox-mcpAdd to Claude Desktop config:
{
"mcpServers": {
"agent-toolbox": {
"command": "agent-toolbox-mcp"
}
}
}See MCP Quickstart Guide for full setup instructions.
pip install langchain-agent-toolboxfrom langchain_agent_toolbox import AgentToolboxSearchTool
search = AgentToolboxSearchTool()
result = search.invoke({"query": "AI agents", "count": 3})pip install llamaindex-agent-toolboxfrom llamaindex_agent_toolbox import AgentToolboxSearchTool
search = AgentToolboxSearchTool()
result = search.call(query="AI agents", count=3)| Example | Description |
|---|---|
| Research Agent | Auto-research any topic → markdown report |
| Website Monitor | Monitor sites with DNS + screenshots + content |
| MCP Quickstart | Claude Desktop / Cursor / Windsurf setup |
| Plan | Price | Calls/month | Rate Limit |
|---|---|---|---|
| Free | $0 | 1,000 | 60/min |
| Pro | $29/mo | 50,000 | 120/min |
| Scale | $99/mo | 500,000 | 300/min |
All plans include all 13 endpoints. No feature gating.
git clone https://github.com/Vincentwei1021/agent-toolbox.git
cd agent-toolbox
npm install && npm run build
cp .env.example .env # Add your config
npm startOr with Docker:
docker build -t agent-toolbox .
docker run -p 3100:3100 --env-file .env agent-toolboxBuilding with Agent Toolbox? Open an issue to get listed here.
MIT © Vincentwei1021