Output Styles in Claude Code: Customising Role, Tone, and Format

ArticleAnthropicResource

Anthropic's documentation details Output Styles in Claude Code, a feature that modifies the system prompt to adjust the assistant's role, tone, and response format throughout a session.

Key points

  1. Output styles alter Claude Code's system prompt (defining role, tone, and format) without altering what Claude knows.
  2. Four built-in styles complement the Default style: Proactive (acts immediately on routine choices without pausing), Concise (direct answers without preamble from v2.1.237), Explanatory (adds educational architectural insights), and Learning (collaborative coding using 'TODO(human)' markers).
  3. Custom output styles are authored as Markdown files with frontmatter in ~/.claude/output-styles, project folders, or bundled in plugins.
  4. The 'keep-coding-instructions' frontmatter parameter controls whether built-in software engineering guidelines remain active (default is false).
  5. Configuration is managed via '/config' in the terminal, the VS Code command menu (v2.1.257+), or the 'outputStyle' key in '.claude/settings.local.json'. The legacy '/output-style' command was removed in v2.1.91.
  6. Modifications take effect following '/clear' or a fresh session; subagents run separate system prompts and do not inherit output styles.

Why it matters

Output styles allow developers and technical teams to repurpose Claude Code for roles beyond standard coding—such as writing assistance, analysis, or structured diagram outputs—while also tailoring verbosity and autonomy levels during daily terminal use.

Analysis

Anthropic clearly delineates output styles from existing configuration layers. Unlike 'CLAUDE.md', which appends codebase guidelines via user messages, output styles rewrite the core system prompt for every turn of the main conversation. A crucial operational detail for custom styles is that engineering guidelines (scoping changes, verification, comments) are discarded by default unless 'keep-coding-instructions: true' is specified. Token usage also varies significantly across styles: Explanatory and Learning increase output tokens, whereas Concise trims output length while preserving safety warnings and error reports.

Mentioned links

Found in

Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • Video

    Video:Alex Sprogis

    Claude Code: Using Output Styles to Tame Overly Complex Claude Opus 5 Responses

    Alex Sprogis demonstrates how to use the built-in 'Output Styles' feature in Claude Code to rein in lengthy, jargon-heavy replies from Claude Opus 5. Using an ELI5 template created by Anthropic developer Lydia Hallie, he explains how to configure, branch-test, and persist custom styles within a project.

    AI & LLMs· Tutorial

  • Link:Anthropic

    Anthropic Claude Code: Agentic Coding in the Terminal and IDEs

    Anthropic introduces Claude Code, an AI coding agent designed for the terminal, IDEs, web, and Slack. The tool maps codebases using agentic search, applies multi-file edits, runs tests, and creates pull requests directly within the developer's environment.

    AI & LLMs· Tool

  • Article:Anthropic

    Extending Claude Code with Skills: Architecture and Best Practices

    Anthropic details the skills system for Claude Code. Developers can encapsulate recurring procedures and workflows inside SKILL.md files built on the open Agent Skills standard, supporting dynamic context injection and subagent delegation.

    AI & LLMs· Tool

  • Article:Anthropic

    How Claude Code Retains Project Context: CLAUDE.md and Auto Memory Explained

    Anthropic outlines the memory architecture of Claude Code. Because every session starts with a clean context window, the CLI tool relies on two complementary systems: user-defined CLAUDE.md files for explicit instructions and an auto memory feature that records learnings from user feedback.

    AI & LLMs· Resource

  • Repository:anthropics/claude-code

    Claude Code: Terminal-Based Agentic Coding Tool

    Claude Code by Anthropic is an agentic coding tool that operates in the terminal. It understands codebases, executes routine development tasks, explains complex code, and handles Git workflows using natural language commands.

    144.1kstarsPython

    AI & LLMs· Tool

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.