curl | bash for AI Agents: One-Command Install for pydantic-deep

The standard Python AI tool install experience:

1. Install Python (which version?)
2. Create a venv
3. pip install
4. ModuleNotFoundError: No module named 'textual'
5. pip install again with correct extras
6. Figure out PATH

Seven steps. Fifteen minutes. That's before you've even seen the tool.

We fixed this for pydantic-deep — the modular agent runtime for Python:

curl -fsSL https://raw.githubusercontent.com/vstorm-co/pydantic-deep/main/install.sh | bash

That's it.

---

What install.sh does

1. Detects whether uv is installed
2. If not: installs uv via the official Astral installer
3. Runs uv tool install "pydantic-deep[cli]" — isolated environment, binary available globally
4. Verifies the install
5. Prints PATH fix instructions if needed

No virtual environment management. No extras guessing. The cli extras group includes everything including textual (which was the original bug — it was missing from the base install).

Self-update

pydantic-deep update

Uses uv tool upgrade if available, falls back to pip. One command to stay current.

Startup notifications

Every invocation checks PyPI for updates silently (2-second timeout, 24-hour cache):

Update available: v0.3.6 → v0.3.7  Run: pydantic-deep update

Never blocks startup.

Why uv?

uv tool install is designed exactly for this use case — isolated tool environments, global binary access, no activation required. Fast, well-maintained, increasingly standard for Python CLI tooling.

Alternatives considered: pipx (slower, needs separate install), Homebrew tap (maintenance overhead), native binary (too brittle for dynamic imports).

---

Full write-up with implementation details: oss.vstorm.co/blog/pydantic-deep-one-command-install-curl-bash

GitHub: github.com/vstorm-co/pydantic-deep

What's the worst install experience you've had with an AI/ML tool?