Llms Txt Generator

LLMs.txt Documentation Generator

This skill generates structured llms.txt documentation files from official library and framework documentation URLs. It creates standardized, LLM-optimized reference files that can be used across conversations for consistent dependency knowledge.

When to Use This Skill

Use this skill when:

• User requests llms.txt file generation for a library, framework, or API
• User provides documentation URL(s) and asks for structured documentation
• User mentions needing dependency documentation in a standardized format
• User asks to "document" or "create docs for" a specific library with a URL

Do not use this skill for:

• General web searches about libraries
• Code generation without documentation needs
• User's own custom documentation (not from official sources)

Documentation Retrieval Protocol

CRITICAL: This skill requires fetching exact URL content, not search results.

Fetching Process

1. Identify the URL fetching capability in the current environment

   • Common tool names: web_fetch, fetch_url, get_page, retrieve_url
   • Look for tools that can retrieve complete page content from exact URLs

2. Use that tool to retrieve the EXACT URL provided by the user

   • Never substitute with web search
   • Never use search results as a proxy for actual documentation

3. If the environment defaults to search or lacks fetching capability:

   • Inform the user: "This environment cannot fetch exact URLs. I need direct page access to generate accurate llms.txt files."
   • Request alternative approach (manual content paste, different environment, etc.)

Validation

Before processing retrieved content, confirm:

• ✓ Content is from the actual documentation page (check page structure, headings)
• ✓ Contains complete information with API references, code examples, or usage patterns
• ✓ URL in response matches the requested URL
• ✗ Content is NOT search engine snippets or results
• ✗ Content is NOT "Page not found", error pages, or truncated previews

If validation fails, inform the user and request the correct URL or clarification.

When URLs Are Missing

If the user requests llms.txt generation without providing URL(s):

1. Ask the user to provide the official documentation URL(s)
2. Explain that official documentation URLs are needed for accurate extraction
3. Example response: "I'll need the official documentation URL for [library]. Please provide the link so I can generate accurate llms.txt documentation."

Do not proceed with search-based alternatives without explicit user consent.

Standard llms.txt Structure

Generate llms.txt files following this template structure (see references/llms-template.md for full template):

Required Sections

1. Header Block

   • Library/Framework name
   • Version (if specified in docs or URL)
   • Official documentation source URL
   • Generation date

2. Overview

   • Brief description (2-3 sentences)
   • Primary use case
   • Key features

3. Installation & Setup

   • Installation commands
   • Basic configuration
   • Peer dependencies (if applicable)

4. Core Concepts

   • Main abstractions or patterns
   • Architecture overview
   • Key terminology

5. API Reference

   • Main functions/methods/components
   • Parameters and return types
   • TypeScript types (if applicable)

6. Common Usage Patterns

   • Practical code examples
   • Frequent use cases
   • Integration patterns

7. Error Handling

   • Common errors
   • Error handling patterns
   • Debugging tips

8. Best Practices

   • Recommended patterns
   • Performance considerations
   • Common pitfalls to avoid

Optional Sections (include if present in documentation)

• Authentication/Authorization
• Advanced Features
• Migration Guides
• Framework-Specific Integration (e.g., React, Vue, Node.js)

Processing Multiple URLs

When user provides multiple URLs for the same library:

1. Fetch all provided URLs
2. Synthesize information into a single coherent llms.txt file
3. Avoid duplication - merge overlapping content
4. Maintain logical section flow from the template

When user provides URLs for different libraries:

1. Generate separate llms.txt files for each library
2. Name files clearly: [library-name].llms.txt
3. Present all files to the user
4. Offer to create them as separate artifacts if requested

Extraction Guidelines

Content Quality

• Prioritize official examples over third-party code
• Include working code snippets that demonstrate actual usage
• Preserve TypeScript types exactly as documented
• Keep examples concise but functional (5-15 lines typically)
• Extract API signatures with parameter types and return types

What to Include

• Installation commands and setup steps
• Core API methods, functions, or components
• Type definitions and interfaces
• Configuration options
• Practical usage examples
• Error handling patterns
• Common integration scenarios

What to Exclude

• Marketing copy and promotional language
• Excessive background/history
• Changelog details
• Contributor information
• Overly verbose explanations (condense to essentials)
• Deprecated APIs (unless specifically requested)

Output Format

File Naming

Use kebab-case: library-name.llms.txt

Examples:

• stripe-api.llms.txt
• react-query.llms.txt
• tanstack-table.llms.txt

File Structure

# Library Name

**Version:** X.X.X
**Source:** [Official Documentation URL]
**Generated:** YYYY-MM-DD

## Overview
[Content]

## Installation & Setup
[Content]

[... remaining sections ...]

Presentation

After generation:

1. Present the complete llms.txt file to the user
2. Confirm the file meets their needs
3. Offer to:
   • Adjust the level of detail
   • Add/remove sections
   • Generate additional llms.txt files for other dependencies
   • Explain how to use the file in their workflow

Context-Aware Generation

Framework Detection

Detect the framework/context from the documentation URL and adjust structure:

• React libraries: Emphasize component usage, hooks, props
• Node.js libraries: Emphasize server-side usage, async patterns, middleware
• API services: Emphasize authentication, endpoints, request/response formats
• Build tools: Emphasize configuration, plugins, optimization

Detail Level

Default to comprehensive documentation that includes:

• Complete API surface
• Multiple usage examples per section
• Type definitions
• Configuration options

If user requests concise format:

• Focus on most common APIs (80/20 rule)
• One example per major feature
• Minimal configuration details
• Skip advanced features

Quality Assurance

Before presenting the final llms.txt file, verify:

1. Completeness: All major API features are documented
2. Accuracy: Information matches the official documentation
3. Functionality: Code examples are syntactically correct
4. Coherence: Sections flow logically
5. Consistency: Formatting is uniform throughout
6. Source attribution: Original documentation URL is referenced

Example Usage Flow

User: "Generate llms.txt for Stripe API from https://stripe.com/docs/api"

Response:

1. Fetch https://stripe.com/docs/api
2. Validate content is official Stripe API documentation
3. Extract according to template structure
4. Generate stripe-api.llms.txt with all sections
5. Present to user with confirmation prompt

User: "Create docs for React Query, here's the link: https://tanstack.com/query/latest"

Response:

1. Fetch the URL
2. Detect this is a React library
3. Emphasize hooks, component patterns
4. Generate react-query.llms.txt
5. Present with React-specific usage examples highlighted