Skip to content
Home › Claude Code Configuration & Workflows

Output Styles ⚠ verify

Claude Code Configuration & Workflows

Definition

Output styles directly modify Claude Code's system prompt (role, tone, output format) and apply to every response — effectively swapping in a different system-prompt persona for the session.

Key points

  • Built-in styles: Default (SW-engineering system prompt), Proactive (executes immediately, prefers action over planning), Explanatory (adds educational "Insights"), Learning (learn-by-doing; shares "Insights" and adds TODO(human) markers).
  • Change via /config → Output style; selection saved to .claude/settings.local.json. Or set directly in a settings file: { "outputStyle": "Explanatory" }.
  • (⚠ verify) The standalone /output-style command was deprecated in v2.1.73 and removed in v2.1.91 — use /config or edit the outputStyle setting.
  • Custom styles = a Markdown file with frontmatter (name, description, keep-coding-instructions [default false], force-for-plugin) at User ~/.claude/output-styles, Project .claude/output-styles, or managed policy. File name becomes the style name unless name is set.

Why it matters for the exam

  • D3 (20% of scored content) tests knowing output styles alter the system prompt (not just tone superficially) and how they are now selected — the removal of /output-style in favor of /config is an exam-worthy version detail.

Common gotchas

  • Output style is part of the system prompt (read once at session start); changes take effect only after /clear or a new session.
  • Custom styles drop Claude Code's built-in SW-engineering instructions unless keep-coding-instructions: true.
  • Do not reach for the standalone /output-style command — it was removed in v2.1.91; use /config.

See also

Sources

Referenced by