Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when building high-performance async Python APIs with FastAPI and Pydantic V2. Invoke to create REST endpoints, define Pydantic models, implement authentication flows, set up async SQLAlchemy database operations, add JWT authentication, build WebSocket endpoints, or generate OpenAPI documentation. Trigger terms: FastAPI, Pydantic, async Python, Python API, REST API Python, SQLAlchemy async, JWT authentication, OpenAPI, Swagger Python.
.claude/skills/jeffallan-fastapi-expert/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 47% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-20 | ✓→✓ | = Same ✓ | 68% | 0% |
Deep expertise in async Python, Pydantic V2, and production-grade API development with FastAPI.
pytest after each endpoint group and verify OpenAPI docs at /docs> Checkpoint after each step: confirm schemas validate correctly, endpoints return expected HTTP status codes, and /docs reflects the intended API surface before proceeding.
Schema + endpoint + dependency injection in one cohesive unit:
python# schemas.py from pydantic import BaseModel, EmailStr, field_validator, model_config class UserCreate(BaseModel): model_config = model_config(str_strip_whitespace=True) email: EmailStr password: str name: str | None = None @field_validator("password") @classmethod def password_strength(cls, v: str) -> str: if len(v) < 8: raise ValueError("Password must be at least 8 characters") return v class UserResponse(BaseModel): model_config = model_config(from_attributes=True) id: int email: EmailStr name: str | None = None
python# routers/users.py from fastapi import APIRouter, Depends, HTTPException, status from sqlalchemy.ext.asyncio import AsyncSession from typing import Annotated from app.database import get_db from app.schemas import UserCreate, UserResponse from app import crud router = APIRouter(prefix="/users", tags=["users"]) DbDep = Annotated[AsyncSession, Depends(get_db)] @router.post("/", response_model=UserResponse, status_code=status.HTTP_201_CREATED) async def create_user(payload: UserCreate, db: DbDep) -> UserResponse: existing = await crud.get_user_by_email(db, payload.email) if existing: raise HTTPException(status_code=status.HTTP_409_CONFLICT, detail="Email already registered") return await crud.create_user(db, payload)
python# crud.py from sqlalchemy import select from sqlalchemy.ext.asyncio import AsyncSession from app.models import User from app.schemas import UserCreate from app.security import hash_password async def get_user_by_email(db: AsyncSession, email: str) -> User | None: result = await db.execute(select(User).where(User.email == email)) return result.scalar_one_or_none() async def create_user(db: AsyncSession, payload: UserCreate) -> User: user = User(email=payload.email, hashed_password=hash_password(payload.password), name=payload.name) db.add(user) await db.commit() await db.refresh(user) return user
python# security.py from datetime import datetime, timedelta, timezone from jose import JWTError, jwt from fastapi import Depends, HTTPException, status from fastapi.security import OAuth2PasswordBearer from typing import Annotated SECRET_KEY = "read-from-env" # use os.environ / settings ALGORITHM = "HS256" oauth2_scheme = OAuth2PasswordBearer(tokenUrl="/auth/token") def create_access_token(subject: str, expires_delta: timedelta = timedelta(minutes=30)) -> str: payload = {"sub": subject, "exp": datetime.now(timezone.utc) + expires_delta} return jwt.encode(payload, SECRET_KEY, algorithm=ALGORITHM) async def get_current_user(token: Annotated[str, Depends(oauth2_scheme)]) -> str: try: data = jwt.decode(token, SECRET_KEY, algorithms=[ALGORITHM]) subject: str | None = data.get("sub") if subject is None: raise ValueError return subject except (JWTError, ValueError): raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="Invalid credentials") CurrentUser = Annotated[str, Depends(get_current_user)]
Load detailed guidance based on context:
| Topic | Reference | Load When | |-------|-----------|-----------| | Pydantic V2 | references/pydantic-v2.md | Creating schemas, validation, model_config | | SQLAlchemy | references/async-sqlalchemy.md | Async database, models, CRUD operations | | Endpoints | references/endpoints-routing.md | APIRouter, dependencies, routing | | Authentication | references/authentication.md | JWT, OAuth2, get_current_user | | Testing | references/testing-async.md | pytest-asyncio, httpx, fixtures | | Django Migration | references/migration-from-django.md | Migrating from Django/DRF to FastAPI |
field_validator, model_validator, model_config)Annotated pattern for dependency injectionX | None instead of Optional[X]@validator, class Config)When implementing FastAPI features, provide:
FastAPI, Pydantic V2, async SQLAlchemy, Alembic migrations, JWT/OAuth2, pytest-asyncio, httpx, BackgroundTasks, WebSockets, dependency injection, OpenAPI/Swagger
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-20 | pass→pass | 13,004 | 13,686 | +5% | 1 | 1 | 0% | 2,783 | 4,672 | +68% | 0 | 0 | — |
case-01 | fail→pass | 14,286 | 8,863 | -38% | 1 | 1 | 0% | 2,643 | 3,397 | +29% | 0 | 0 | — |
case-02 | fail→pass | 16,856 | 17,426 | +3% | 1 | 1 | 0% | 3,241 | 4,760 | +47% | 0 | 0 | — |
case-03 | fail→pass | 21,422 | 12,893 | -40% | 1 | 1 | 0% | 4,809 | 4,485 | -7% | 0 | 0 | — |
case-04 | pass→pass | 6,625 | 6,167 | -7% | 1 | 1 | 0% | 1,212 | 2,664 | +120% | 0 | 0 | — |
case-05 | pass→pass | 12,054 | 12,594 | +4% | 1 | 1 | 0% | 2,266 | 2,984 | +32% | 0 | 0 | — |
case-06 | pass→pass | 5,609 | 7,002 | +25% | 1 | 1 | 0% | 1,183 | 3,140 | +165% | 0 | 0 | — |
case-07 | pass→pass | 7,112 | 7,600 | +7% | 1 | 1 | 0% | 1,412 | 3,019 | +114% | 0 | 0 | — |
case-08 | pass→pass | 34,134 | 15,035 | -56% | 1 | 1 | 0% | 2,718 | 4,554 | +68% | 0 | 0 | — |
case-09 | pass→pass | 9,989 | 5,633 | -44% | 1 | 1 | 0% | 1,406 | 2,533 | +80% | 0 | 0 | — |
case-10 | pass→pass | 8,390 | 10,603 | +26% | 1 | 1 | 0% | 1,481 | 3,060 | +107% | 0 | 0 | — |
case-11 | fail→pass | 13,975 | 7,584 | -46% | 1 | 1 | 0% | 2,506 | 2,921 | +17% | 0 | 0 | — |
case-12 | pass→pass | 10,108 | 8,291 | -18% | 1 | 1 | 0% | 1,856 | 3,372 | +82% | 0 | 0 | — |
case-13 | pass→pass | 7,978 | 10,192 | +28% | 1 | 1 | 0% | 1,475 | 3,439 | +133% | 0 | 0 | — |
case-14 | pass→pass | 8,101 | 7,360 | -9% | 1 | 1 | 0% | 1,577 | 2,904 | +84% | 0 | 0 | — |
case-15 | pass→pass | 11,892 | 7,591 | -36% | 1 | 1 | 0% | 1,985 | 3,046 | +53% | 0 | 0 | — |
case-16 | pass→pass | 10,125 | 19,897 | +97% | 1 | 1 | 0% | 1,897 | 3,974 | +109% | 0 | 0 | — |
case-17 | pass→pass | 8,327 | 9,400 | +13% | 1 | 1 | 0% | 1,596 | 3,249 | +104% | 0 | 0 | — |
case-18 | pass→pass | 9,732 | 9,701 | -0% | 1 | 1 | 0% | 1,763 | 3,194 | +81% | 0 | 0 | — |
case-19 | pass→pass | 6,134 | 6,614 | +8% | 1 | 1 | 0% | 1,153 | 2,987 | +159% | 0 | 0 | — |
case-21 | pass→pass | 13,063 | 19,804 | +52% | 1 | 1 | 0% | 2,762 | 4,234 | +53% | 0 | 0 | — |
case-22 | pass→pass | 11,674 | 8,040 | -31% | 1 | 1 | 0% | 2,328 | 2,941 | +26% | 0 | 0 | — |
case-23 | pass→pass | 12,682 | 8,164 | -36% | 1 | 1 | 0% | 2,714 | 2,999 | +11% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +17 percentage points is the difference between those two pass rates over the 23 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.