Container deployments require consistent image, networking, configuration, and runtime decisions. This skill provides reusable patterns for Docker, Kubernetes, Helm, and ECS/Fargate.

Claude Code hooks are powerful, but unsafe hook scripts can leak data or block work unexpectedly. This skill explains hook events, configuration patterns, and safer script templates for controlled automation.

Teams need repeatable Claude Code runs without an interactive terminal. This skill explains headless CLI usage, structured output, CI patterns, and tool controls.

Long-running dev servers and watchers can lose output or block agent work. This skill gives tmux patterns for starting, monitoring, restarting, and isolating those processes.

Local Tilt environments can hide failing resources, stale logs, and Docker Compose health problems. This skill guides Claude, Codex, and Claude Code through status checks, logs, and Tiltfile fixes.

Parallel local project copies can collide on ports, data, cookies, and Docker resources. This skill guides creation or review of a gen-env command that isolates each instance.

Manual PR setup can miss changed files, tests, or repository conventions. This skill guides a Japanese-language workflow for review, commit, push, and GitHub PR creation.