Doc AI
Provides Amazon Alexa+ with conversational access to current weather, forecasts, alerts, and historical weather data from stored and simulated IoT sources.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Doc AIWhat's the weather like in Austin today?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Doc AI
Conversational weather agent for the Amazon Developer Hackathon — Alexa+ track only.
Doc AI absorbs public weather APIs plus simulated IoT sensors, stores normalized readings in SQLite, and exposes them through a self-hosted MCP server (spec 2025-11-25, Streamable HTTP) so Alexa+ can talk about current conditions, forecasts, alerts, and history in short spoken sentences.
This is not a rename of DOC Watch dock-advice tools. Doc AI is a first-release ingest/store/chat product with historical + IoT + chat_summary.
One-command start
cd doc-ai
npm install
npm run devThen:
Health: http://127.0.0.1:8790/health
Alexa+ sim: http://127.0.0.1:8790/
MCP:
POSTandGEThttp://127.0.0.1:8790/mcp
Tests: npm test
Related MCP server: MCP Weather Server
MCP tools
Tool | What it does |
| Live Open-Meteo (seed fallback) |
| Daily forecast, |
| NWS |
| Date-range query against stored readings |
| Simulated IoT nodes |
| Spoken briefing, no markdown |
Data layer
Open-Meteo
api.open-meteo.com+ archivearchive-api.open-meteo.comNWS
api.weather.govSimulated IoT probes (labeled simulation — not a live worldwide camera mesh)
SQLite (
data/doc-ai.sqlite) seeded fromdata/seed.jsonso the demo works offline
Folders
server/ MCP Streamable HTTP entry (POST + GET /mcp, Origin 403)
data/ ingest, normalize, SQLite, seed, IoT sim
agent/ voice formatter + tool orchestrator
deploy/ Dockerfile companion, App Runner script
docs/ product-feedback, friction log, Devpost paste, demo script
web/ simulated Alexa+ chatDeploy (AWS App Runner)
docker build -t doc-ai .
# or
bash deploy/deploy-apprunner.shRequest AWS Builder credits if needed. Localhost bind is the default for the 3-minute demo.
License
MIT. Public GitHub intended: https://github.com/Scrum723/doc-ai
If submitting private instead, share with testing@devpost.com and Amazon reviewers: chris-trag, knmeiss, giolaq, anishamalde, mosesroth, emersonsklar.
Prior work
DOC / WFAD / SWAFT / DOC Watch are related weather-agent lineage. Doc AI is a new ingest-store-MCP surface (historical, IoT list, chat_summary, two-API SQLite), not a reskin of the locked DOC Watch dock-tool list.
Primary track: Alexa+ only.
This server cannot be deployed
Maintenance
Related MCP Connectors
Global weather via Open-Meteo: forecast, historical, marine, air quality, geocoding, elevation.
Provide real-time and forecast weather information for locations in the United States using natura…
US weather & geo for AI agents: forecasts, alerts, earthquakes, elevation, geocoding. No keys.
US weather & geo for AI agents: forecasts, alerts, earthquakes, elevation, geocoding. No keys.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables users to get weather alerts and forecasts for U.S. locations using the National Weather Service API, with support for natural language queries including fun conversational tones.MIT
- AlicenseNot gradedqualityDmaintenanceProvides real-time weather information and multi-day forecasts using the Open-Meteo API through a robust, enterprise-grade architecture. It enables AI assistants to retrieve current conditions, 7-day forecasts, and contextual weather data for any location.7 npmMIT
- AlicenseNot gradedqualityCmaintenanceEnables natural language queries about Ambient Weather personal weather station data, including current conditions and device info.MIT
- FlicenseNot gradedqualityBmaintenanceExposes live weather data from Open-Meteo and NWS, providing tools to explore forecasts and list active severe weather alerts. It renders an embedded interactive widget for the Athena platform, with no API keys required.-