README.md

Fastango is a uv-first CLI for generating FastAPI projects with polished terminal flows, safe AI-assisted planning, selectable templates, and a growing integration catalog.

It helps teams start a serious FastAPI codebase quickly without copying random boilerplate, hardcoding secrets, or letting an AI write unchecked files.

Why Fastango

Need	What Fastango Gives You
Start a FastAPI app quickly	simple and mvc templates with settings, routes, tests, README, and llms.txt.
Add common infrastructure	A searchable catalog for auth, billing, databases, queues, storage, observability, AI, deployment, and dev tools.
Generate from product ideas	fastango generate "a starter MVP" maps prompts to supported templates, skills, presets, and integrations.
Keep generation safe	The generator only uses Fastango-supported tools and validates provider suggestions before writing files.
Work like modern Python teams	Generated projects use uv, Ruff, pytest, typed settings, and clear next-step commands.

Highlights

• Branded interactive terminal playground when you run uvx fastango.
• Manual project creation with repeatable flags for CI and documentation.
• Constrained AI generator for prompt-first FastAPI scaffolding.
• Live model discovery for Anthropic and OpenAI API keys.
• Curated presets for starter APIs, SaaS apps, AI APIs, data APIs, and production setups.
• Generated llms.txt so AI assistants understand the project conventions.
• Validation for integration requirements, conflicts, unsupported tools, and unsafe provider output.

Install

uv tool install fastango

During local development:

uv sync --all-groups
uv run fastango --help

Quick Start

Run Fastango with no subcommand to open the terminal playground:

uvx fastango

You can also launch the playground explicitly:

fastango playground

Create a project without prompts:

fastango create billing-api \
  --style mvc \
  --python 3.12 \
  --preset saas \
  --integration openapi \
  --integration authx \
  --integration stripe \
  --with-docker \
  --no-interactive

Run the generated app:

cd billing-api
uv sync
cp .env.example .env
uv run fastapi dev app/main.py

Generate From A Prompt

Fastango can infer a constrained FastAPI scaffold from a natural-language prompt. It does not write arbitrary LLM code directly to disk. It converts your prompt into a typed plan made from supported templates, skills, presets, and integrations.

fastango generate "a starter MVP with auth, billing, email, analytics and secure webhooks"
fastango generate "AI SaaS with vector search" --dry-run --json

Optional provider enrichment is available when configured:

fastango generate "marketplace MVP with billing and uploads" --provider anthropic
fastango generate "AI SaaS with vector search" --provider openai --model gpt-4.1-mini

Provider output is validated against the same supported catalog before any files are written. Unsupported requests are kept as not_generated notes in the preview.

Model Discovery

If the matching API key is set, Fastango asks the provider which models your key can access. If no key is available, it falls back to a curated offline list.

fastango models
fastango models --provider anthropic
fastango models --provider openai --json
fastango models --provider openai --static

Provider	Environment Variable	Model Source
Anthropic	ANTHROPIC_API_KEY	Live /v1/models discovery with curated fallback.
OpenAI	OPENAI_API_KEY	Live /v1/models discovery with curated fallback.

Templates

Template	Best For	Includes
simple	Small APIs, internal services, demos, first FastAPI apps.	app/main.py, routes, schemas, settings, tests, README, .env.example, llms.txt.
mvc	SaaS apps, production APIs, larger teams, integration-heavy projects.	API routes, core settings, services, repositories, schemas, tests, README, .env.example, llms.txt.

Presets

Preset	Purpose	Typical Stack
api-starter	A clean FastAPI baseline.	OpenAPI, tests, CORS, Ruff, pre-commit.
saas	Product-ready SaaS API.	AuthX, Stripe, Postgres, Redis, Resend, PostHog, Sentry, Docker.
ai-api	AI or RAG-ready backend.	OpenAI, Anthropic, pgvector, Postgres, Redis, tests, Docker.
data-api	Data-heavy API with workers.	Postgres, Alembic, Redis, Celery, Prometheus, OpenTelemetry.
production	Production hardening.	Docker, GitHub Actions, security headers, Sentry, Prometheus, health checks.

Integration Catalog

List, filter, search, and inspect integrations:

uv run fastango integrations
uv run fastango integrations --category database
uv run fastango integrations --search vector
uv run fastango integrations --presets
uv run fastango integrations --json

Category	Examples
Auth and security	AuthX, FastAPI Users, OAuth, JWT, CORS, rate limiting, CSRF, security headers, roles, API keys.
Databases and ORM	Postgres, SQLite, MySQL, MongoDB, SQLModel, Tortoise, Alembic, Supabase, pgvector.
Cache and jobs	Redis, Celery, Dramatiq, ARQ, RQ, APScheduler, FastAPI background tasks.
SaaS and payments	Stripe, Paddle, Polar, Lemon Squeezy, subscriptions, customer portal, Resend, PostHog, Sentry.
Storage	S3, Cloudflare R2, Google Cloud Storage, local files, Pillow, uploads.
AI and search	OpenAI, Anthropic, Ollama, LangChain, LlamaIndex, Qdrant, Pinecone, Weaviate, Elasticsearch.
Observability	OpenTelemetry, Prometheus, Structlog, Logfire, health checks, audit logs.
API protocols	OpenAPI, GraphQL, WebSockets, SSE, signed webhooks, versioning.
Deployment and dev tools	Docker, Compose overlays, Kubernetes, GitHub Actions, pre-commit, Ruff, mypy, pytest, Dependabot.
Code-template packs	Production hooks, SaaS routes, workers, realtime routes, storage services, and AI/RAG service boundaries.

Generation Skills

Skills are internal, allowlisted generation capabilities. They map product intents to supported Fastango templates and integrations.

Skill	What It Builds
saas-mvp	Auth, teams, subscriptions, billing provider, email, analytics, monitoring.
secure-api	CORS, security headers, rate limiting, signed webhooks, API keys, secure settings.
ai-api	LLM provider, vector store, RAG-ready services, cache, background jobs.
marketplace	Users, teams, payments, uploads, webhooks, audit logs.
crud-api	Database, CRUD routes, pagination, filters, tests.
production-api	Docker, GitHub Actions, health checks, observability, dependency hygiene.

Safety Model

Rule	Why It Matters
No arbitrary raw code from providers	LLMs can suggest supported IDs, but Fastango writes files only through templates and integration hooks.
Secrets stay in settings and .env.example	Generated code avoids hardcoded tokens, API keys, and webhook secrets.
Registry validation before writes	Presets, aliases, integration requirements, and conflicts are resolved before the filesystem is touched.
Unsupported tech becomes a note	Requests for unsupported frameworks are shown in previews instead of silently generated.

Development

uv sync --all-groups
uv run pytest
uv run ruff check .
uv run mypy fastango

Status

Fastango is early and evolving quickly. The current focus is a great terminal experience, a reliable integration catalog, and safe prompt-to-FastAPI scaffolding.