multiAI Summary Pending

generating-output-styles

Creates custom output styles for Claude Code that modify system prompts and behavior. Use when the user asks to create output styles, customize Claude's response format, generate output-style files, or mentions output style configuration.

231 stars

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/generating-output-styles/SKILL.md --create-dirs "https://raw.githubusercontent.com/aiskillstore/marketplace/main/skills/emz1998/generating-output-styles/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/generating-output-styles/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How generating-output-styles Compares

Feature / Agentgenerating-output-stylesStandard Approach
Platform SupportmultiLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

Creates custom output styles for Claude Code that modify system prompts and behavior. Use when the user asks to create output styles, customize Claude's response format, generate output-style files, or mentions output style configuration.

Which AI agents support this skill?

This skill is compatible with multi.

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

SKILL.md Source

# Generating Output Styles

Creates custom output styles for Claude Code following Anthropic's standards and best practices.

## Context

Output styles modify Claude Code's system prompt to adapt its behavior beyond software engineering. They directly affect the main agent loop and control response tone, structure, and approach.

**Key capabilities:**
- Generate valid output style markdown files with frontmatter
- Apply proper naming and description standards
- Structure instructions following progressive disclosure
- Configure keep-coding-instructions appropriately
- Place files in correct locations (user vs project level)

## Workflow

### Phase 1: Gather Requirements

1. Determine the output style purpose and target behavior
2. Identify if coding instructions should be retained
3. Choose placement: `~/.claude/output-styles/` (user) or `.claude/output-styles/` (project)
4. Reference output style documentation at `.claude/skills/output-style/references/output-style-docs.md`

### Phase 2: Create Output Style File

1. **Craft frontmatter:**
   - `name`: Clear, descriptive name (shown in UI)
   - `description`: Brief explanation for users
   - `keep-coding-instructions`: true/false (default: false)

2. **Write system prompt modifications:**
   - Start with role definition
   - Define specific behaviors and approaches
   - Include formatting guidelines
   - Add any special instructions or constraints
   - Keep instructions concise and actionable

3. **Structure content:**
   - Use markdown headers for organization
   - Provide concrete examples when helpful
   - Avoid contradicting core Claude Code functionality
   - Focus on additive instructions, not restrictions

### Phase 3: Validate and Save

1. Verify YAML frontmatter syntax (opening/closing ---)
2. Ensure description is user-friendly and clear
3. Confirm keep-coding-instructions matches requirements
4. Save to appropriate location
5. Test activation with `/output-style` command

## Implementation Strategy

**File format:**
```markdown
---
name: Style Name
description: User-facing description of what this style does
keep-coding-instructions: false
---

# Custom Style Instructions

You are an interactive CLI tool that helps users with software engineering tasks. [Custom instructions...]

## Specific Behaviors

[Define how the assistant should behave in this style...]
```

**Naming guidelines:**
- Use title case for `name` field
- Keep names concise but descriptive
- Avoid technical jargon in names
- Examples: "Explanatory", "Learning", "Technical Writer"

**Description guidelines:**
- Write for end users, not developers
- Explain what the style does, not how
- Keep under 200 characters
- Focus on benefits and use cases

**keep-coding-instructions:**
- `false` (default): Removes software engineering instructions, clean slate
- `true`: Retains coding guidance, adds modifications on top

## Constraints

- Must include valid YAML frontmatter with all required fields
- System prompt modifications should complement, not contradict, core Claude Code
- Cannot override tool permissions or security restrictions
- File must use `.md` extension
- Must be placed in recognized output-styles directory
- Description shown in UI, so must be user-friendly

## Success Criteria

- Valid markdown file with proper YAML frontmatter
- Name and description follow guidelines
- keep-coding-instructions set appropriately
- System prompt instructions are clear and actionable
- File saved to correct location
- Can be activated via `/output-style` command
- Produces expected behavioral changes when active

## Examples

**Example 1: Technical Writer Style**

```markdown
---
name: Technical Writer
description: Produces detailed documentation with comprehensive explanations
keep-coding-instructions: true
---

# Technical Writer Instructions

You are an interactive CLI tool that helps users with software engineering tasks.

## Documentation Approach

- Provide comprehensive explanations for all code changes
- Include detailed comments in code
- Document edge cases and assumptions
- Create thorough README sections when relevant
- Explain trade-offs in implementation decisions

## Formatting

- Use clear section headers
- Include code examples with explanations
- Add inline comments for complex logic
- Structure responses with intro, body, conclusion
```

**Example 2: Minimalist Style**

```markdown
---
name: Minimalist
description: Provides concise responses with minimal explanation
keep-coding-instructions: true
---

# Minimalist Instructions

You are an interactive CLI tool that helps users with software engineering tasks.

## Communication Style

- Keep responses under 5 lines when possible
- Use code without explanatory prose
- Omit obvious explanations
- Respond with direct answers
- Only elaborate when explicitly asked
```

## Quick Reference

**Common output style patterns:**

1. **Educational styles**: Set `keep-coding-instructions: true`, add explanatory guidance
2. **Specialized domain styles**: Set `keep-coding-instructions: false`, define new role
3. **Formatting styles**: Set `keep-coding-instructions: true`, add response structure rules
4. **Tone modifiers**: Set `keep-coding-instructions: true`, adjust communication style

**Testing workflow:**
1. Create output style file
2. Run `/output-style` to verify it appears in menu
3. Activate the style
4. Test with representative tasks
5. Iterate based on behavior
6. Verify style persists in `.claude/settings.local.json`

## Related Documentation

- Output Style Docs: `.claude/skills/output-style/references/output-style-docs.md`
- Output Style Template: `.claude/templates/output-style.md` (if exists)
- Settings Documentation: Claude Code settings reference