Tools & Features

Key concepts

• › Prefer dedicated tools to Bash — Read/Edit/Write beat cat/sed/echo on UX and on safety
• › Parallelise independent tool calls — multiple tool uses in one assistant turn run concurrently
• › Delegate to a sub-agent to protect context — large research drops belong in an Explore agent, not the main thread
• › TodoWrite is for tracking — not for narration. Mark tasks done as you finish them, not in batch
• › Hooks fire on every event of their type — keep them fast or you slow the whole session down

Reference template

// Reference: the built-in tool families
1. File tools          (Read, Edit, Write — read-before-edit invariant)
2. Shell               (Bash — persistent cwd, parallel calls, background jobs)
3. Search              (Grep, Glob — fast targeted lookups)
4. Delegation          (Task / Agent — sub-agents in parallel or background)
5. Web                 (WebFetch, WebSearch — cached content, search results)
6. Tracking            (TodoWrite — task list with status)
7. Orchestration       (Skill, slash commands — wrappers over the above)
8. Events              (Hooks — PreToolUse, PostToolUse, UserPromptSubmit, Stop)

Adapt to your problem; the structure is the load-bearing part.

Common pitfalls

• ⚠ Reaching for Bash when a dedicated tool exists — cat for reading, sed for editing, echo > file for writing are all worse than the dedicated tools
• ⚠ Sequencing independent calls — three reads chained costs three round-trips when one parallel batch would do
• ⚠ Running an unbounded grep / find from / — large filesystem scans waste tokens and time
• ⚠ Using the Task tool to do work that should be in the main thread — delegation is for parallelism and context-protection, not for hiding work

Related topics

Items (11)

• File Tools — Read, Edit, Write

  The three primary file tools. When to use each, the read-before-edit invariant, and why Edit is preferred over Write for existing files.

  Feature Foundational
• The Bash Tool

  Running shell commands, persistent working directory, parallel execution, background jobs, and the sandbox model.

  Feature Foundational
• Search Tools — Grep and Glob

  Fast targeted search across the repository. When to prefer Grep over Bash-ripgrep, when to delegate broader searches to an Explore agent.

  Feature Foundational
• The Task Tool — Background Work

  Running long operations in the background, getting notified on completion, and the no-polling rule that makes this fast.

  Feature Intermediate
• WebFetch and WebSearch

  Pulling content from a URL or searching the web from inside a session. Caching, content extraction, and where the limits sit.

  Feature Foundational
• TodoWrite and Task Management

  The built-in task list. When to use it, when not to, and how it interacts with sub-agent task tracking.

  Feature Foundational
• Slash Commands

  Built-in slash commands (/help, /clear, /loop, /ultrareview, /schedule), and the ones you'll use most as a daily driver.

  Feature Foundational
• Skills — User-Invocable Workflows

  Bundled prompt + tool + instruction packages that Claude Code can invoke. How they differ from sub-agents and custom slash commands.

  Feature Intermediate
• Sub-Agents

  Specialised agents (Explore, Plan, claude-code-guide, general-purpose). Parallel and isolated execution, when to delegate, and context-window protection.

  Feature Intermediate
• Hooks Overview

  Event-driven shell hooks that fire on tool calls and prompt submission. The four event types and where each one is the right tool.

  Feature Intermediate
• The Status Line

  Customising the always-visible status line — model, branch, token usage, custom indicators. Useful for long sessions.

  Feature Intermediate