بنقرة واحدة
chdb-lambda-microvm-demo
يحتوي chdb-lambda-microvm-demo على 6 من skills المجمعة من nklmish، مع تغطية مهنية على مستوى المستودع وصفحات skill داخل الموقع.
Skills في هذا المستودع
Deploy the NYC Taxi agent to AWS AgentCore Runtime in the current account — provision full CDK infra (VPC, ECR, IAM, S3 Files, Monitoring, CICD), AgentCore Memory, build+push the image, and register/poll the runtime. Use when asked to "deploy to agentcore", "run on agentcore runtime", or deploy the agent to AWS.
Deploy the NYC Taxi agent to AWS Lambda MicroVMs in the current account — package the app as a MicroVM image with chDB baked in and lifecycle hooks, run a Firecracker-isolated MicroVM with its own private chDB, and verify over the dedicated HTTPS endpoint. Use when asked to "deploy to lambda microvms", "run on microvms", "deploy the microvm demo", or show the chDB × Lambda MicroVMs reference architecture.
Deploy and verify the S3 Files NFS mount demo on EC2 — runs the agent container on a Graviton EC2 host that mounts the S3 Files filesystem at /mnt/noaa-gsod via the amazon-efs-utils helper so weather queries use file() instead of the S3 fallback. Use when asked to demo or verify the S3 Files mount, since AgentCore Runtime itself cannot mount it.
One command for "deploy everything to AWS and let me use it from my browser". Deploys the full stack to AWS AgentCore (full data, Memory, S3 Files mount demo), then runs the chat UI locally pointed at the deployed runtime so the browser talks to the AWS agent. Use when asked to "deploy to prod", "deploy to AWS and run the UI", "deploy everything", or run the full cloud stack with a local browser.
Launch and drive the NYC Taxi Analytics Agent locally for a fast smoke test — bake a tiny chDB sample, start the FastAPI/uvicorn server, and hit /chat. Use when asked to run, start, or verify this app on a dev machine without the full Finch/AWS container flow.
Run the NYC Taxi Analytics Agent locally in PROD MODE — provision AgentCore Memory in the current AWS account (idempotent), then launch the FastAPI app with IS_PROD=true and memory enabled, and drive a /chat that proves memory retrieval + write. Use when asked to "run in prod", "run prod mode", or start the app with AgentCore Memory.