llms-txt

/llms-txt

Generate an llms.txt file (and optionally llms-full.txt) following the llmstxt.org specification. This provides AI coding assistants and search engines with a structured index of your project's documentation.

Behaviour

1. Load the llms-txt skill for the specification and generation patterns
2. Load the doc-standards rule for description quality
3. Read the project manifest (package.json, pyproject.toml, etc.) for name and description
4. Scan the repository for documentation files:
   • Core: README.md, docs/, API reference
   • Guides: docs/guides/
   • Examples: examples/
   • Supporting: CONTRIBUTING.md, CHANGELOG.md, SECURITY.md, CODE_OF_CONDUCT.md, ROADMAP.md, LICENSE
5. Write benefit-focused descriptions for each file (not just file names)
6. Assemble llms.txt following the spec:
   • H1 from project name
   • Blockquote from manifest description or README first paragraph
   • H2 sections grouping docs by category
   • ## Optional for supporting files
7. If full argument: concatenate all referenced files into llms-full.txt

Output Files

File	Content	When
llms.txt	Index with relative links and benefit-focused descriptions	Always
llms-full.txt	Concatenated Markdown of all referenced docs	Only with full argument

Description Quality

Every file annotation must be benefit-focused:

Good: [Getting Started](./docs/guides/getting-started.md): Install, configure, and deploy your first worker in under 5 minutes

Bad: [Getting Started](./docs/guides/getting-started.md): Getting started guide

Use at least 3 different benefit categories across the file (Time saved, Confidence gained, Pain avoided, Capability unlocked, Cost reduced).

Arguments

• No arguments: generate llms.txt only for the current project
• full: generate both llms.txt and llms-full.txt
• Path argument: generate for a specific project directory