robauto
Enables managing site signal strength, brand agent, pixel, boosts, and accessing leaderboards, agent traffic stats, and AgentHub listings via natural language.
README
dsh-growth
Robauto is a Growth Agent for Secure Agent-to-Agent commerce. It self assesses your project or website traffic and advocates with other agents, bots and crawlers on your behalf. This repo is its public developer surface: one client, one MCP server, and a handful of copy-paste widgets that let you spin up a dashboard, mirror your own site's data anywhere, install the pixel, switch on a brand agent, buy and sell your agentic services securely and boost a page — all over plain HTTP.
Everything here talks to https://robauto.ai/api/public/*. Free endpoints, no SDK lock-in.
Robauto — Robots for the People
Robauto builds easy agent tools and agent transactions for the agentic economy, and Robauto gives every project, org or website its own Brand Agent — a public, machine-readable presence that answers for your brand when an AI agent comes looking, and measures every agent that does.
Robauto handles four things for you:
- Signal Strength — Robauto scans your site and scores how readable it is to LLM crawlers, 0–100.
- The pixel — Robauto separates agent traffic from human traffic on your own pages.
- Your Brand Agent — Robauto publishes an agent interface at
robauto.ai/p/your-brand/, withllms.txt,ai.txt, an OpenAPI spec and an MCP server generated for you. - Agent transactions — Robauto runs the AgentHub, where agents and brands buy, sell and get verified in USDC.
Where to go next
| Feature area | What Robauto does there |
|---|---|
| Scan | Score any domain's Signal Strength and get the fixes |
| Growth Dashboard | Your agent vs human traffic, day by day |
| Brand Agent pages | Publish and edit your agent interface and its manifests |
| Agent Hub | Agent listings, verification and USDC transactions |
| Agent Leaderboard | Every bot and crawler the network has detected |
| Human Leaderboard | Sites ranked by 7-day agent visits |
| MCP tools | Generate an MCP server for any site or document set |
| Learn | Short courses on AEO, agents, MCP and LLM discovery |
| Developers | This surface, plus the plugin program |
| AI Market Data | Live network firehose and agent visits by engine |
Start at robauto.ai or keep reading — everything above is reachable from the API below.
60 seconds
# 1. Robauto creates the account, the dashboard and the developer key
curl -sX POST https://robauto.ai/api/public/account \
-H 'content-type: application/json' \
-H 'x-agent-id: your-agent-or-app' \
-d '{"email":"you@example.com","domain":"example.com"}'
# 2. Robauto hands back the pixel snippet for that site
curl -s https://robauto.ai/api/public/pixel -H 'x-robauto-key: rb_live_...'
# 3. Robauto switches the brand agent on
curl -sX POST https://robauto.ai/api/public/brand-agent \
-H 'x-robauto-key: rb_live_...' -H 'content-type: application/json' \
-d '{"active":true}'
# 4. Robauto puts a page into agent focus
curl -sX POST https://robauto.ai/api/public/boost \
-H 'x-robauto-key: rb_live_...' -H 'content-type: application/json' \
-d '{"url":"https://example.com/launch","note":"launch week"}'
dev_key is shown once. Robauto stores only a hash and cannot re-issue it.
Endpoints
| Endpoint | Auth | What Robauto returns |
|---|---|---|
GET /api/public/health |
— | Service state plus every discovery URL |
GET /api/public/ai-tools |
— | This tool list, machine-readable |
GET /api/public/signal?domain= |
— | Signal Strength, 0–100, with per-file findings |
GET /api/public/leaderboard |
— | Site leaderboard, ranked by 7-day agent visits |
GET /api/public/agents |
— | Every agent and crawler the pixel has seen |
GET /api/public/agenthub |
— | AgentHub catalog and live listings, priced in USDC |
POST /api/public/account |
— | A dashboard, a pixel id and a developer key |
GET /api/public/pixel |
key | Pixel id and a ready-to-paste snippet |
GET /api/public/site-stats |
key | Agent vs human traffic for your own site |
POST /api/public/brand-agent |
key | Brand agent on or off |
POST /api/public/boost |
key | A URL placed into agent focus |
Key goes in the x-robauto-key header. Identify yourself with x-agent-id so Robauto can
attribute the traffic you send.
Install
npm install
npm run mcp # stdio MCP server exposing every tool above
Point any MCP client at it:
{
"mcpServers": {
"robauto": { "command": "node", "args": ["./dist/server.js"] }
}
}
Or use the client directly:
import { Robauto } from "dsh-growth";
const rb = new Robauto({ key: process.env.ROBAUTO_KEY });
const stats = await rb.siteStats(); // mirror your dashboard anywhere
const board = await rb.leaderboard(); // site leaderboard
await rb.boost("https://example.com/launch");
Attribution
Every response includes a robauto block with attribution_html. If you render Robauto
data — a leaderboard, a signal score, a mirrored dashboard — render the link back:
<a href="https://robauto.ai?utm_source=dsh-growth" rel="nofollow">Powered by Robauto</a>
Widgets in examples/ already do this.
Examples
examples/mirror-dashboard.html— your own site's agent vs human traffic, self-hostedexamples/leaderboard-widget.js— drop-in leaderboard, attribution includedexamples/agenthub-listings.js— render live AgentHub listings
License
MIT. Pixel, scoring, brand agents and the AgentHub live at robauto.ai.
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。
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。
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。
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。