enferno

Lean Flask framework built for coding agents: Vue 3 + Vuetify with no build step, production-ready auth (2FA, passkeys, OAuth), AGENTS.md, an agent skill and a read-only MCP server. No webpack, no node_modules, just Python.

Automatically checked · GitHub checked 2026-11-14

enferno

Overview

Enferno 🔥

Checks License: MIT

Docs · Tutorials

A Flask framework built for coding agents. You describe the app; your agent builds it on a stack it already understands, and a test suite tells it when it got something wrong.

Plain Python, Vue 3 and Vuetify with no build step, production auth out of the box. No webpack, no node_modules, nothing for the agent to configure.

Start

Paste this into Claude Code, Codex, Cursor or any coding agent:

Clone https://github.com/level09/enferno into ./myapp, follow its AGENTS.md
to set it up, create an admin user, start it on port 5001 and run the tests.

Or by hand:

git clone https://github.com/level09/enferno.git myapp && cd myapp
./setup.sh                    # Installs deps, generates .env with secrets
uv run flask create-db        # Create the database
uv run flask install          # Create an admin user
uv run flask run              # http://localhost:5000

setup.sh never prompts when there is no terminal, so agents and CI run it with flags: --full adds Redis and Celery, --docker configures the Compose stack.

Open video on GitHub

Then just ask

Add a Product model with name, price and category, an admin page to manage it,
and a migration.

Let users export the activity log as CSV.

Add a status column to users and show it as a chip in the users table.

How your agent knows Enferno

What it gives the agentWhen
AGENTS.mdStack, layout, commands, routes and hard rules, in about 100 linesEvery session, read automatically
enferno-dev skillFull patterns for models, APIs, Vue pages, dialogs, migrations and upgradesLoaded only when a task needs it
flask mcpLive routes, schema with row counts, migration revision and audit log, read-onlyWhen the agent needs the real state, not the docs
/llms.txtA short description of your app for LLM toolsEdit it to describe what you build

The skill lives in .agents/skills/, the shared location for Codex, Cursor, Copilot and Gemini CLI, and .claude/skills links to it for Claude Code. npx skills add level09/enferno installs it in another project. MCP is opt-in: uv sync --extra dev --extra mcp, and .mcp.json lets MCP-aware agents offer to connect.

How your agent checks its work

Agents write code fast; they also forget migrations and leave routes unprotected. uv run pytest catches the mistakes they actually make:

  • Every protected route rejects anonymous users, and every admin route rejects non-admins, including routes the agent just added.
  • Models match migrations: change a model without flask db migrate and the suite fails.
  • CSRF is enforced on the admin API, and a valid token still works.
  • Auth flows: login, wrong password, POST-only logout, audited admin CRUD.
  • setup.sh in every mode, without touching your .env.

The app tests run on in-memory SQLite in about two seconds; the setup.sh tests add about ten. CI runs them on PostgreSQL across Python 3.11 to 3.14. Set TEST_DATABASE_URI to use PostgreSQL locally; it must be a disposable database, because the suite drops every table.

AGENTS.md also gives the agent the rules that tests cannot: Tabler icons only, POST-only logout, migrations reviewed by hand, and a real browser check after a frontend upgrade.

What's in the box

  • Frontend: Vue 3, Vuetify 4 and axios, vendored and pinned by vendor.sh. No build step
  • Auth: login, registration, password recovery, TOTP 2FA, passkeys (WebAuthn), recovery codes, Google and GitHub OAuth
  • Admin: users, roles and an audit log of admin actions
  • Database: SQLAlchemy 2 with Alembic migrations. SQLite by default, PostgreSQL when you want it
  • UI: collapsible sidebar, dark mode, Tabler icons, data tables, dialogs and notifications
  • Optional: Redis sessions and Celery (--extra full), MCP (--extra mcp), uWSGI (--extra wsgi)

Requires Python 3.11+ and uv.

Deploy

Docker, with PostgreSQL, Redis, Celery and nginx:

./setup.sh --docker
docker compose up --build

Compose requires DB_PASSWORD and REDIS_PASSWORD, which setup.sh --docker generates. For an existing installation, keep the credentials its database and Redis already use. Database and Redis ports stay internal to the Compose network.

Any Ubuntu VPS, with Caddy and automatic SSL, via Ignite:

curl -sSL https://raw.githubusercontent.com/level09/ignite/main/ignite.sh | sudo DOMAIN=your-domain.com bash

Configuration

setup.sh writes .env with generated secrets. The values you are most likely to change:

FLASK_DEBUG=1                    # 0 in production

# PostgreSQL (optional, SQLite works by default)
# SQLALCHEMY_DATABASE_URI=postgresql://user:pass@localhost/dbname

# Background tasks (optional)
# REDIS_URL=redis://localhost:6379/1
# CELERY_BROKER_URL=redis://localhost:6379/2

Full guides: docs.enferno.io

Related

  • ReadyKit adds Stripe, workspaces and teams on top of Enferno.
  • STK is the same philosophy on Quart (async Flask) with native WebSockets.

License

MIT