The Shortform Guide

Tip

This guide covers the essentials of ECC in a condensed format. For the full deep-dive, see the Longform Guide.

Skills and Commands

Skills operate like rules, scoped to certain workflows. They’re shorthand to prompts when you need to execute a particular workflow.

After a long session of coding, you want to clean out dead code? Run /refactor-clean. Need testing? /tdd, /e2e, /test-coverage.

Commands are skills executed via slash commands. They overlap but are stored differently:

• Skills: ~/.claude/skills/ - broader workflow definitions
• Commands: ~/.claude/commands/ - quick executable prompts

Hooks

Hooks are trigger-based automations that fire on specific events. Unlike skills, they’re tied to tool calls and lifecycle events.

Hook Types:

1. PreToolUse - Before a tool executes (validation, reminders)
2. PostToolUse - After a tool finishes (formatting, feedback loops)
3. UserPromptSubmit - When you send a message
4. Stop - When Claude finishes responding
5. PreCompact - Before context compaction
6. Notification - Permission requests

Agents

Subagents are specialized AI instances that handle specific tasks. ECC ships with 28 agents organized by function:

• Code Reviewers - Language-specific review (TypeScript, Python, Go, Rust, Java, Kotlin, C++, Flutter)
• Build Resolvers - Fix build errors (generic, Go, Java, Kotlin, Rust, C++, PyTorch)
• Workflow Agents - Planner, architect, TDD guide, refactorer, doc updater

Each agent declares its tools, model tier (Opus/Sonnet/Haiku), and specialization in YAML frontmatter.

MCP Servers

Model Context Protocol servers extend Claude’s capabilities with external integrations. ECC includes 24 server configurations for GitHub, databases, deployment platforms, browser automation, and more.

Best practice: Keep fewer than 10 MCP servers active to preserve context window.

Getting More Out of ECC

• Token optimization: Use model routing to match task complexity with the right model tier
• Memory persistence: Hooks auto-save context across sessions
• Continuous learning: Extract patterns from sessions into reusable skills
• Verification loops: Run evals to ensure quality

For detailed coverage of these topics, see the Longform Guide.