Installation guide — learning_ai_devops_tools Purpose This repository contains interactive, safe installers and helpers to install CLI tools (Claude Code, OpenAI Codex, Antigravity agy, Devin, GitHub Copilot) on WSL/Ubuntu, macOS, and Windows. Files of interest - install_clis_wsl.sh — interactive WSL installer (WSL/Ubuntu). Preview and confirm before running remote installers. - make_symlinks_wsl.sh — creates /usr/local/bin symlinks (requires sudo) - run_installers.sh — cross-platform wrapper to run installers from WSL or show instructions - run_installers.ps1 — Windows PowerShell wrapper to run WSL or show Windows-native steps - cli-install-report.md — generated report of installs (example) Quick start (WSL/Ubuntu) 1. Open WSL (Ubuntu) shell. 2. cd /mnt/d/SANDBOX/mygh/learning_ai_devops_tools 3. Ensure scripts use LF and are executable: sudo apt-get update && sudo apt-get install -y dos2unix dos2unix install_clis_wsl.sh run_installers.sh make_symlinks_wsl.sh || true chmod +x install_clis_wsl.sh run_installers.sh make_symlinks_wsl.sh 4. Run the interactive installer (will preview each remote installer and ask confirmation): bash -i ./install_clis_wsl.sh Quick start (Windows PowerShell with WSL) - From PowerShell run (recommended): wsl bash -ic "cd /mnt/d/SANDBOX/mygh/learning_ai_devops_tools && dos2unix install_clis_wsl.sh || true && bash -i ./install_clis_wsl.sh" Quick start (macOS) - Inspect installers first. macOS support is similar to Linux; use the run_installers.sh wrapper to list commands. Do NOT pipe unknown scripts to shell without review. Security and safety - All remote installers are previewed before execution. - No secrets or API keys are written to shell profiles. - Auth steps are left interactive (use the tool's login commands). Developer notes - Use .gitattributes to enforce LF endings on shell scripts across platforms. - To reproduce: run the scripts from a fresh WSL Ubuntu session and follow interactive prompts. If you want, run './run_installers.sh' to get an interactive cross-platform flow.