Skip to content

Quickstart

This walkthrough takes you from an empty project to a FastAPI service with metrics, health checks, CORS, Swagger, and structured logging, wired up in a few lines. The same pattern applies to Litestar, FastStream, FastMCP, and framework-free services.

1. Install

Install lite-bootstrap with the FastAPI extras. fastapi-all pulls in every supported instrument:

uv add "lite-bootstrap[fastapi-all]"

pip install "lite-bootstrap[fastapi-all]"

poetry add "lite-bootstrap[fastapi-all]"

Want only some instruments? See the extras matrix in Installation.

2. Bootstrap your application

Create main.py:

from fastapi import FastAPI
from lite_bootstrap import FastAPIConfig, FastAPIBootstrapper

config = FastAPIConfig(
    service_name="my-service",
    service_version="1.0.0",
    service_environment="local",
    prometheus_metrics_path="/metrics",
    health_checks_enabled=True,
    health_checks_path="/health",
    cors_allowed_origins=["*"],
    swagger_offline_docs=True,
)

app: FastAPI = FastAPIBootstrapper(config).bootstrap()


@app.get("/")
async def root() -> dict[str, str]:
    return {"hello": "world"}

bootstrap() returns a ready FastAPI app with every configured instrument attached.

3. Run it

uvicorn main:app --reload

4. What you get

With the config above, your service now exposes:

  • Prometheus metrics at /metrics, for your scraper.
  • A health check at /health, for your orchestrator's readiness probe.
  • Swagger UI at /docs, with offline assets so it renders without a CDN.
  • CORS headers for the configured origins.
  • Structured logging via structlog, on by default (opt out with logging_enabled=False).

5. Enable more

Logging, health checks, Prometheus and Swagger are on by default whenever their extras are installed. OpenTelemetry, Sentry, Pyroscope and CORS stay off until you set their config fields; unset (or empty) fields are skipped automatically.

config = FastAPIConfig(
    service_name="my-service",
    service_version="1.0.0",
    service_environment="local",
    # tracing — spans exported to your OTLP endpoint
    opentelemetry_endpoint="http://localhost:4317",
    # error reporting
    sentry_dsn="https://examplePublicKey@o0.ingest.sentry.io/0",
    # continuous profiling (needs the `pyroscope` extra)
    pyroscope_endpoint="http://localhost:4040",
)

To see which instruments were enabled or skipped at startup, inspect the bootstrapper. skipped_instruments lists only instruments skipped by configuration; a configured instrument whose extra is not installed is skipped with an InstrumentDependencyMissingWarning instead.

bootstrapper = FastAPIBootstrapper(config)
app = bootstrapper.bootstrap()

print(bootstrapper.build_summary())  # human-readable summary
for cls, reason in bootstrapper.skipped_instruments:
    print(f"skipped {cls.__name__}: {reason}")

Next steps

  • Configuration: every config field and per-instrument option.
  • Integrations: Litestar, FastStream, FastMCP, and framework-free services.