#automation
6 articles
Autonomous Agents: From Task Lists to Reasoning Systems
Your agents run tasks in parallel. This article teaches them to think — planning, reflection, error recovery, and cost-aware decisions. Includes a working code reviewer agent blueprint.
Claude Code on Autopilot: Four Hooks You Install Once and Never Think About Again
Auto-commit when Claude finishes a task. Auto-lint after every file change. Load project context at session start. Get notified when long tasks finish. Four pre-built automations — copy, paste, done.
Fifteen Claude Code Features Most People Never Use
Session teleport, mobile access, loops, hooks, dispatch, voice input, worktrees, and more. Every hidden feature the creator of Claude Code uses daily — explained.
Build Your Own Skill Pack: The Complete Guide to Claude Code Skills
Six skills you can install today. Multi-agent code audits, adversarial debates, creative councils, autonomous loops, and more. Each one teaches a pattern you can adapt to anything you build.
What Are Claude Code Skills (And Why They Change Everything)
Slash commands are one-shot prompts. Skills are entire folders — with scripts, assets, hooks, and memory. They turn Claude from a tool you talk to into a collaborator who knows your workflow.
Claude Learned to Work Without You
On April 8, Anthropic gave Claude Code servers of its own. Agents no longer need your laptop open, your desktop app running, or you in the room. Here's what that actually means for creatives.