Goal Workflows with /goal
Turn a prompt into a durable objective. Claude works across turns until a small fast model confirms the condition is met. Pairs with auto mode for fully unattended work.
Advanced Techniques
Claude Code becomes a fundamentally different tool once you move past basic prompting. Hooks give you deterministic guardrails over a probabilistic system. Custom commands encode your team’s workflows into repeatable actions. The memory hierarchy lets you shape Claude’s behavior at the organization, project, and personal level. And headless mode turns your terminal agent into a CI/CD pipeline worker.
This section covers the techniques that separate casual users from teams shipping production code with Claude Code every day.
What You Will Find Here
Section titled “What You Will Find Here”Dynamic Workflows & ultracode
Orchestrate tens to hundreds of subagents from a script Claude writes and you can rerun. Turn on ultracode and Claude decides when a task warrants a workflow. Research preview as of v2.1.154.
Agent View: claude agents
The dashboard for background sessions. Dispatch new agents, peek at progress, attach when one needs you. Research preview as of v2.1.139.
Slash Commands Mastery
Every built-in command, keyboard shortcut, and session management pattern. The interactive reference you keep open in a second terminal.
Hooks and Automation
The hook lifecycle from SessionStart to SessionEnd. Block dangerous commands, auto-format on save, inject context, and build deterministic guardrails around a probabilistic agent.
Custom Commands
Create reusable slash commands that encode your team’s best practices. From simple prompt templates to full agent-powered workflows.
Memory System
The full CLAUDE.md hierarchy, auto memory, project rules, and imports. Control what Claude knows at every scope level.
Enterprise Integration
Corporate SSO, managed settings, compliance policies, and team-wide configuration patterns for organizations deploying Claude Code at scale.
Proxy Configuration
HTTP proxies, VPN tunnels, corporate firewalls, and certificate management. Get Claude Code working behind any network infrastructure.
LLM Gateway Setup
Route Claude Code through AWS Bedrock, Google Vertex AI, Azure, LiteLLM, and custom API gateways. Control where your tokens flow.
Dev Container Workflows
Isolated, reproducible development environments with Claude Code pre-configured. Codespaces, devcontainers, and remote development patterns.
GitHub Actions
Run Claude Code in CI/CD. Automated code reviews, PR generation, issue triage, and scheduled maintenance — all driven by the Claude Code Action.
Monitoring and Costs
OpenTelemetry metrics, cost tracking, token optimization, and team budget management. Know exactly where your money goes.
Where to Start
Section titled “Where to Start”If you are new to advanced Claude Code usage, start with Slash Commands Mastery to build fluency with the interactive REPL, then move to Hooks and Automation to understand how deterministic control layers work. Memory System is essential reading for any team deploying Claude Code across multiple developers.
If you want to leverage the May 2026 autonomous-work features, read Goal Workflows, Agent View, and Dynamic Workflows together — dispatch verifiable goals into background sessions, check on the dashboard when each row turns green, and reach for a workflow when a task needs tens to hundreds of agents orchestrated from a script Claude writes.
If you are deploying Claude Code for a team, read Enterprise Integration and Monitoring and Costs first, then work through Proxy Configuration or LLM Gateway Setup based on your infrastructure requirements.