You need to ship reliable software, but management only gave you 2 weeks.
No time for health checks, proper testing, or clean architecture. Just enough time for duct tape and hope.
What if you could go from idea to working prototype in the time it takes to grab coffee?
Ship FastAPI apps that grow with you.
Aegis Stack scaffolds complete FastAPI applications with auth, payments, workers, AI, and a built-in control plane. Add what you need, remove what you don't, update when the framework improves.
uvx aegis-stack init my-api && cd my-api && make serveMore examples
# Add user authentication out of the box
uvx aegis-stack init user-app --services auth
# Add background processing + scheduling
uvx aegis-stack init task-processor --components scheduler,worker
# Everything wired up at init
uvx aegis-stack init full-app --services auth,payment,comms --components worker,schedulerCLI in 9 languages: English, German, Spanish, French, Japanese, Korean, Russian, Simplified Chinese, Traditional Chinese. Use
aegis --lang <code>or setAEGIS_LANG.
Installation alternatives: See the Installation Guide for uv tool install, pip install, and development setup.
Components are infrastructure pieces (database, workers, scheduler, cache). Services are business capabilities (auth, AI, payments, comms).
Don't worry about what you pick today. Add anything later with a single command, remove what you outgrow, no rework required.
| Component | What you get | |
|---|---|---|
| Backend | FastAPI + lifecycle hooks | |
| CLI | Typer, first-class system interface | |
| Frontend | Flet, ships with the Overseer system dashboard | |
| Cache | Redis for caching, sessions, pub/sub | |
| Database | Postgres or SQLite + SQLModel ORM | |
| Inference | Local AI models via Ollama | |
| Ingress | Traefik v3 reverse proxy with TLS | |
| Observability | Pydantic Logfire tracing + metrics + logging | |
| Scheduler | APScheduler with persistent jobs | |
| Worker | Pluggable Arq, Taskiq, or Dramatiq |
| Service | What you get | |
|---|---|---|
| AI | Conversational agents, RAG, model catalog, TTS and STT (PydanticAI / LangChain across 7 providers) | |
| Auth | JWT cookie sessions, GitHub/Google sign-in, RBAC, multi-tenant Organizations | |
| Blog | Markdown publishing with drafts, tags, and an Overseer editor | |
| Comms | Transactional email (Resend) + SMS / voice (Twilio) | |
| Insights | Adoption metrics across GitHub, PyPI, Plausible, Reddit | |
| Payments | Stripe checkout, subscriptions, refunds, disputes |
| Capability | What you get | |
|---|---|---|
| Deploy CLI | One-command deploys to any VPS over SSH (rsync + Docker), no PaaS lock-in | |
| Server Setup | aegis deploy-setup provisions Ubuntu, Debian, or Fedora boxes (Docker + firewall) |
|
| Backups & Rollback | pg_dump before every deploy, retention policy, automatic rollback on failed health checks |
|
| TLS / HTTPS | Let's Encrypt via the Traefik ingress component, zero config when a domain is set |
Components compose into capabilities you didn't have to build:
graph LR
Auth[Auth] --> A[User-specific AI conversations]
AI[AI] --> A
AI --> B[Persistent history &<br/>token analytics]
DB[(Database)] --> B
AI --> C[Background AI pipelines]
Worker[Worker] --> C
Scheduler[Scheduler] --> D[Persistent job scheduling]
DB --> D
Your choices aren't permanent. Start with what you need today, add when requirements change, remove what you outgrow.
# Monday: Ship MVP
aegis init my-api
# Week 3: Add scheduled reports
aegis add scheduler --project-path ./my-api
# Month 2: Need async workers
aegis add worker --project-path ./my-api
# Month 6: Scheduler not needed
aegis remove scheduler --project-path ./my-api
# Stay current with template improvements
aegis update| Starter | Add Later? | Remove Later? | Git Conflicts? |
|---|---|---|---|
| Others | Locked at init | Manual deletion | High risk |
| Aegis Stack | One command | One command | Auto-handled |
Most starters lock you in at init. Aegis Stack doesn't. See Evolving Your Stack for the complete guide.
Live Demo: sector-7g.dev/dashboard
Overseer is the embedded control plane that ships with every Aegis Stack project.
- Live health of every component and service in one view
- Worker queues, scheduled jobs, recent runs
- Database schema, tables, and migration state
- AI token usage and conversation history
- Auth sessions and user activity
- No external tooling, no vendor integrations, no setup
The Aegis CLI is a first-class interface to your running system. Not just a health check, but a full inspection layer.
- Component-aware commands for every running subsystem
- Inspect worker queues, scheduler runs, database state
- Query AI usage, auth sessions, service configuration
- Same data Overseer sees, terminal-native
- Built into every generated project, no extra installs
When the AI service is enabled, Aegis exposes an additional interface: Illiana, a conversational operator over your running system.
- Ask questions in plain language about live system state
- Backed by live telemetry from Overseer and the CLI
- Optional RAG over your codebase for code-aware answers
- Opt-in: only available when the AI service is on
- Nothing in the stack depends on her being there
- Overseer - Built-in system dashboard
- Deployment - Deploy with backups, rollback, and health checks
- CLI Reference - Complete command reference
- Evolving Your Stack - Add/remove components as needs change
- Technology Stack - Battle-tested technology choices
- About - The philosophy and vision behind Aegis Stack
No reinventing the wheel. Just the tools you already know, pre-configured and ready to compose.
Aegis Stack respects your expertise. No custom abstractions or proprietary patterns to learn. Pick your components, get a production-ready foundation, and build your way.
Aegis gets out of your way so you can get started.





