Tower

0.3.61 verified Fri May 01 auth: no python

Tower is a CLI and runtime environment for managing and running agentic workflows, currently at version 0.3.61. It provides a framework for defining, orchestrating, and executing multi-step AI agent tasks. Release cadence is irregular as it is in early development.

pip install tower

Common errors

error ImportError: cannot import name 'Tower' from 'tower' ↓

cause The library was installed incorrectly or an old version is cached.

fix

Run 'pip install --upgrade tower' and verify the installation with 'pip show tower'.

error AttributeError: module 'tower' has no attribute 'run' ↓

cause The 'run' method was removed in version 0.3.50; use 'execute' instead.

fix

Replace 'tower.run(...)' with 'tower.execute(...)'.

error tower.exceptions.AuthenticationError: Invalid API key ↓

cause The TOWER_API_KEY environment variable is missing or incorrect.

fix

Set TOWER_API_KEY to a valid API key from your Tower account.

Warnings

breaking The Tower API is unstable and may change without notice. Breaking changes are common in minor version bumps. ↓

fix Pin your dependencies and test upgrades thoroughly.

deprecated The 'tower run' CLI command has been deprecated in favor of 'tower execute'. ↓

fix Replace 'tower run' with 'tower execute' in scripts.

gotcha Running Tower without setting TOWER_API_KEY will cause silent fallback to a demo mode with limited functionality. ↓

fix Always set TOWER_API_KEY environment variable before importing tower.

Imports

Tower
```
from tower import Tower
```
Primary class for initializing a Tower session.

Quickstart

Initialize Tower and run a simple workflow. Requires TOWER_API_KEY environment variable.

from tower import Tower
import os

# Initialize Tower with API key from environment
api_key = os.environ.get('TOWER_API_KEY', '')
tower = Tower(api_key=api_key)

# Example: run a simple workflow
result = tower.run("Hello, Tower!")
print(result)