NodeAPI

NodeAPI

Machine-native GIS processing API for AI agents and developers. Convert, reproject, validate, repair, buffer, clip, dissolve, and tile vector geodata across 25 endpoints. Pay-per-use USDC on Solana Mainnet ($0.01/op). No accounts, no API keys. Remote MCP SSE.

Category
访问服务器

README

Node API

Version: 0.5.2
Status: Phase 3 — Public
Live at: nodeapi.ai | API Docs | MCP

Machine-native spatial data processing for AI agents and developers. 25 endpoints. Pay $0.01 USDC per operation on Solana. No accounts, no API keys, no subscriptions.


Quick Start

# 1. Send a request — get a 402 with payment details
curl -X POST https://nodeapi.ai/v1/validate \
  -F "file=@parcels.geojson"

# → 402: { "recipient": "D8m8C9amSawdqSgEdXWkMZ3M86qsVQuDJswG1wZYkezP", "amount_usd": "0.01", ... }

# 2. Send 0.01 USDC to the recipient on Solana Mainnet
#    Get the transaction signature from your wallet

# 3. Retry with the transaction signature
curl -X POST https://nodeapi.ai/v1/validate \
  -F "file=@parcels.geojson" \
  -H "X-PAYMENT: <your_solana_tx_signature>"

# → 200: { "total_features": 1204, "valid_count": 1204, ... }

Payment

Every paid endpoint uses Solana Pay with USDC on Solana Mainnet.

Protocol Solana Pay
Network Solana Mainnet
Asset USDC (EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v)
Price $0.01 flat per operation
Recipient D8m8C9amSawdqSgEdXWkMZ3M86qsVQuDJswG1wZYkezP

Flow:

  1. Call any endpoint → receive 402 with Solana Pay URL + recipient
  2. Send 0.01 USDC on Solana Mainnet (include memo: nodeapi:<operation>)
  3. Retry with X-PAYMENT: <transaction_signature>
  4. Receive processed data

Endpoints

All endpoints available at /v1/<path> (recommended) and /<path> (legacy, still supported).

Info (free)

Method Path Description
GET /v1/health Service health
GET /v1/pricing Full pricing list

Format & Inspection

Method Path Description
POST /v1/convert Convert between GeoJSON, Shapefile, GeoPackage, KML
POST /v1/reproject Project or reproject to any EPSG (auto-detects assign vs transform)
POST /v1/validate Validate geometry — returns JSON report
POST /v1/repair Repair invalid geometry (make_valid)
POST /v1/schema Extract field schema, CRS, bbox, feature count
POST /v1/dxf Extract geometry from DXF/CAD files

Geoprocessing (single input)

Method Path Description
POST /v1/clip Clip to bounding box or polygon mask
POST /v1/buffer Buffer features by distance in meters
POST /v1/erase Delete all features, preserve empty schema
POST /v1/dissolve Dissolve features by attribute field
POST /v1/feature-to-point Convert geometries to centroid points
POST /v1/feature-to-line Extract polygon boundaries as lines
POST /v1/feature-to-polygon Convert closed lines to polygons
POST /v1/multipart-to-singlepart Explode multipart geometries
POST /v1/add-field Add a new attribute column

Geoprocessing (two inputs)

Method Path Description
POST /v1/union Combine all features from two layers
POST /v1/intersect Spatial intersection of two layers
POST /v1/difference Layer A minus overlap with layer B
POST /v1/append Append features from layer B into layer A's schema
POST /v1/merge Merge two layers, preserving all fields
POST /v1/spatial-join Join attributes by spatial relationship

Tiles

Method Path Description
POST /v1/vectorize Generate .mbtiles vector tile package (tippecanoe)

Jobs

Method Path Description
GET /v1/jobs/{id} Poll async job status
GET /v1/jobs/{id}/download Download completed job result

Supported input formats: GeoJSON, Shapefile (.zip), GeoPackage (.gpkg), KML, GDB (read), DXF


MCP Integration

Node API exposes all 25 endpoints as MCP tools — compatible with Claude, GPT-4o, and any MCP-enabled agent.

Remote (no install required)

https://nodeapi.ai/mcp/sse

Connect any MCP client to this SSE endpoint. No authentication needed.

Claude Desktop (local)

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "node-api": {
      "command": "/path/to/venv/bin/python",
      "args": ["-m", "app.mcp_server"],
      "cwd": "/path/to/node-api"
    }
  }
}

22 MCP tools: convert, reproject, validate, repair, schema, clip, dxf, buffer, union, intersect, difference, erase, dissolve, feature-to-point, feature-to-line, feature-to-polygon, multipart-to-singlepart, add-field, append, merge, spatial-join, pricing


Webhooks

For async jobs (DXF conversion, any large file), pass a webhook_url to receive a POST on completion instead of polling.

curl -X POST https://nodeapi.ai/v1/dxf \
  -F "file=@site-plan.dxf" \
  -F "output_format=geojson" \
  -F "webhook_url=https://your-agent.com/callback" \
  -H "X-PAYMENT: <tx_signature>"

Webhook payload:

{
  "job_id": "abc-123",
  "operation": "dxf",
  "status": "done",
  "result_url": "https://nodeapi.ai/v1/jobs/abc-123/download",
  "result_filename": "output.geojson",
  "size_bytes": 48291
}

Signed with X-Webhook-Signature: sha256=<HMAC> for verification.


Architecture

  • FastAPI — async, OpenAPI spec at /docs
  • GeoPandas + GDAL — 200+ vector format support
  • Shapely 2.x — geometry operations
  • pyproj — CRS reprojection
  • tippecanoe 2.79 — vector tile generation
  • PostgreSQL — operations log + job queue
  • Solana Pay — USDC micropayments on Solana Mainnet
  • MCP SDK — stdio (local) + SSE (remote) transport
  • Cloudflare Tunnelnodeapi.ai → API server

Dev Setup

git clone https://github.com/eianray/node-api
cd node-api
python3 -m venv venv
venv/bin/pip install -r requirements.txt
cp .env.example .env
# Leave SOLANA_WALLET_ADDRESS=YOUR_SOLANA_WALLET_ADDRESS for dev mode (bypasses payment)
venv/bin/uvicorn app.main:app --reload --port 8100

Roadmap

  • Phase 1 ✅ — Convert, reproject, validate, repair, schema, clip
  • Phase 2 ✅ — DXF/CAD, buffer, topology (union/intersect/difference), remote MCP SSE
  • Phase 3 ✅ (Mar 4 2026) — 11 new geoprocessing endpoints, Solana Pay, webhooks, vector tiles, API versioning, 22 MCP tools
  • Phase 4 — Hetzner migration, CDN caching, batch API, agent directory listings

nodeapi.ai — spatial processing infrastructure for the agent economy.

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选