•
Install Claude Code, open a practice folder, run one terminal agent task, and review edits—evening-friendly path using official Anthropic docs.
Claude Code is Anthropic’s agentic coding tool: it reads your codebase, edits files, runs commands, and hooks into your existing tools. You can use it from the terminal, IDE extensions, a desktop app, or the web. Official docs: code.claude.com/docs/en.
For after-work Python, the terminal CLI is often the lightest path. You cd into a small folder, type claude, describe a change, watch it work, then inspect git (or a manual diff) before you close the laptop.
People who already open a terminal after work and want an agent that can fix a script, add a test, or explain a file—without rebuilding their whole IDE setup. You need a Claude subscription or Anthropic Console account for most surfaces (see the docs for current options). Keep practice work in a personal folder.
Comparing evening agent tools? Start with Top 10 AI agent tools for after-work Python.
Follow the install block on code.claude.com/docs/en. On macOS/Linux/WSL the recommended native install is:
curl -fsSL https://claude.ai/install.sh | bash
Windows users should use the PowerShell or CMD installers shown on that page (do not mix shells). Homebrew (brew install --cask claude-code) and WinGet are alternatives; those channels may need manual upgrades.
mkdir -p ~/evening-python/claude-lab && cd ~/evening-python/claude-lab
printf '%s\n' 'def add(a, b):' ' return a + b' > calc.py
claude
Sign in on first run. Prefer a narrow project directory so the agent is not staring at your entire home folder.
In the Claude Code session, try:
Add type hints and a
if __name__ == "__main__":block that printsadd(2, 3). Do not create new files. Keep the function pure.
Claude Code can plan, edit, and run commands. Stop it if it starts refactoring beyond the ask.
Exit or pause the agent and run:
python calc.py
Expect 5. Read the file. If it added pytest or packaging you did not request, revert and tighten the prompt.
The docs describe CLAUDE.md at the project root for persistent instructions (style, “stdlib only,” “ask before new deps”). For after-work labs, three bullet rules are enough.
Ask Claude Code to add multiply(a, b) with type hints and print both add and multiply results from __main__. Then ask it to write a three-line docstring only—no other changes. Practice saying no to scope creep.
Claude Code multiplies the Python you already understand. Build that base with SoloQuant’s after-work course Intro to Python Programming—short lessons designed for evenings, so your prompts stay precise.