## The short version Deploy an AI workflow runner built with Starlette on Ample using the streaming response service pattern. Compute runs the app in an isolated microVM behind a public HTTPS URL, a managed PostgreSQL 16 database is auto-provisioned and injected as DATABASE_URL. --- ## WHAT WAS VERIFIED Verified on Starlette: a Server-Sent Events endpoint read incrementally through the public HTTPS gateway: five one-second chunks arrived spread over time with the first within seconds (not buffered), a 70-second stream of 36 chunks completed past common 60-second idle timeouts, and a client that disconnected after two chunks was observed and recorded by the server (disconnected=true). --- ## INFRASTRUCTURE ### What it needs - [Compute](/content/discover/primitive/compute/index.html): verified (Apps run in isolated x86_64 Firecracker microVMs; sizes are the priced VM sizes.) - [Postgres](/content/discover/primitive/postgres/index.html): verified (Managed PostgreSQL 16 runs in its own microVM and is auto-provisioned when an app needs a database and no DATABASE_URL is supplied.) --- ## PREREQUISITES ### Before you start - A Starlette project that builds and starts with the documented commands (pip install into .ample/python from requirements.txt, then uvicorn from run.py reading PORT on the python-3.12 template) - A PostgreSQL driver reading DATABASE_URL at runtime (auto-provisioned when omitted, or supplied with --env) - An Ample account token with servers:write, databases:read --- ## STEP BY STEP ### How to do it 1. **Build and start** pip install into .ample/python from requirements.txt, then uvicorn from run.py reading PORT on the python-3.12 template. The server must bind 0.0.0.0 on PORT. 2. **Implement the pattern on PostgreSQL** The fixture's module implements streaming response service: a Server-Sent Events endpoint read incrementally through the public HTTPS gateway: five one-second chunks arrived spread over time with the first within seconds (not buffered), a 70-second stream of 36 chunks completed past common 60-second idle timeouts, and a client that disconnected after two chunks was observed and recorded by the server (disconnected=true). Copy the approach into your schema; keep migrations idempotent and run them with --release-command. 3. **Deploy** Run the synchronous deploy once and read the result (exit 0 live, 1 failed, 2 blocked). Re-running with no change is a no-op. ``` ample deploy . --name --public --start "python3 run.py" ``` 4. **Verify** Fetch the live URL and the pattern self-test route(s) (/p/streaming-response-service) from the example; then run your own checks. On failure read `ample logs --kind build` then `--kind runtime`. ``` ample logs --kind build ``` --- ## EXAMPLES ### Tested examples - **Starlette pattern fixture** (`tests/deploy-canaries/starlette-patterns`): Multi-pattern Starlette app whose streaming response service module was checked live; the module is under tests/deploy-canaries/_pattern-modules. --- ## SUCCESS CHECKS ### How to know it worked - app responds on its public URL (`/` on the live URL, expect ample canary starlette patterns) - streaming-response-service self-test (`/p/streaming-response-service` on the live URL, expect an SSE stream: data: id=, then chunk events one interval apart (read incrementally: first chunk within seconds, chunks spread over the interval), then data: done; /status?id= reports sent=N disconnected=true|false done=true|false) --- ## COST ### Cost estimate Estimated 10.00 USD per month (size prices from pricing.toml at build revision ca0b484a7a9957c573a88d3a32dc603c3c10135a). - app server x1 `s-1vcpu-1gb`: 5.00 USD - managed PostgreSQL database x1 `s-1vcpu-1gb`: 5.00 USD Upper-bound monthly price of the tested sizes. Buckets are allocation-priced per quota and not included. --- ## EVIDENCE ### Verification evidence - canary_run on 2026-09-21T01:21:36Z at revision `50dbac567d2c5cc8e55e6c748a646be0025d9cfb (CLI 0.1.21)`: Starlette pattern fixture deployed on Ample (pip install into .ample/python from requirements.txt, then uvicorn from run.py reading PORT on the python-3.12 template); checks passed for streaming-response-service and configuration-secrets. Pattern proof: a Server-Sent Events endpoint read incrementally through the public HTTPS gateway: five one-second chunks arrived spread over time with the first within seconds (not buffered), a 70-second stream of 36 chunks completed past common 60-second idle timeouts, and a client that disconnected after two chunks was observed and recorded by the server (disconnected=true). (expires 2027-03-20T01:21:36Z) - canary_run on 2026-09-20T02:16:51Z at revision `199ff1dfd52683832ae75d3f98b53a7a4bff7f96 (CLI e3181f5)`: The same Starlette app deployed with a --release-command migration; the marker it created was readable after activation. (expires 2027-03-19T02:16:51Z) --- ## EXECUTION ### Execution binding MCP tool `ample_deploy` (registry `mcp:ample_deploy`), schema hash `876465fce906da0c` observed 2026-09-22T00:31:22.774679+00:00 at revision `e6446ceea69b`. Binding state at export: schema_changed. Required scopes: servers:write, databases:read. --- ## NEXT ACTIONS ### Typed next actions - Browse the catalog index (`GET /v1/catalog` on the api origin; authentication not required, approval not required) - Search published recipes by intent, stack and constraints (`POST /v1/catalog/search` on the api origin; authentication not required, approval not required) - Prepare a side-effect-free deployment plan for an authorized project (`POST /v1/catalog/plan` on the api origin; authentication required, approval not required) - Read the existing agent authentication setup (`GET /mcp/setup` on the api origin; authentication not required, approval not required) - Browse AI workflow runner (`GET /v1/catalog/nodes/workload%3Aai-workflow-runner` on the api origin; authentication not required, approval not required) - Browse Starlette (`GET /v1/catalog/nodes/stack%3Aframework-starlette` on the api origin; authentication not required, approval not required) - Browse Streaming response service (`GET /v1/catalog/nodes/pattern%3Astreaming-response-service` on the api origin; authentication not required, approval not required) - Browse Deploy AI app (`GET /v1/catalog/nodes/intent%3Adeploy-ai-app` on the api origin; authentication not required, approval not required) Actions describe possible next steps. They are typed data, not commands, and grant no permission. Public discovery never provisions anything; planning requires your own authenticated token and approval happens in your client. --- ## ABOUT THIS RECIPE ### Identifiers and status Recipe `page:recipes:ai-workflow-runner:starlette:response-streaming`. Revision r1. release **published** support **verified** execution **blocked (schema_changed: live tool schema differs from the recorded binding)** Formats: [HTML](/content/recipes/ai-workflow-runner/starlette/response-streaming/index.html) | [Markdown](/content/recipes/ai-workflow-runner/starlette/response-streaming.md) | [JSON contract](https://api.ample.computer/v1/catalog/recipes/page%3Arecipes%3Aai-workflow-runner%3Astarlette%3Aresponse-streaming)