# How We Got LLMs to Recommend Our Open Source Library — Christopher Burns, Inth

## Executive summary

The talk details how modern developer experience primitives must adapt for consumption by Large Language Models (LLMs) and autonomous agents. The core argument is that optimizing documentation requires moving beyond traditional SEO methods to focus on agent-facing artifacts. Key strategies include manually curating `LLMs.txt`, serving bundled markdown files, and ensuring the repository structure (`node_modules`) contains necessary context for coding agents.

## Key takeaways

- Agent Inbound is the New Standard: Inbound traffic sources are shifting from human discovery to LLMs (e.g., Claude, ChatGPT) recommending libraries like C15T. This requires optimizing for 'agent primitives' rather than just developer experience.
- Manual Optimization Beats Automation: For critical files like `LLMs.txt`, manually writing the content is significantly more effective than generating it automatically; 'forty good lines beat a thousand lines of noise.'
- Bundling Context in Node Modules: Coding agents rarely visit live documentation sites. To improve discoverability, developers should ship bundled markdown documents and an `AGENTS.md` file directly within the package's `node_modules` directory.

## Technical details

- Agent-Facing Documentation Pipeline: The solution involves a framework-neutral docs pipeline (Lead Type) that processes `.mdx` files to generate optimized agent experience assets. This is not a magic SEO tool, but an architectural optimization.
- Serving Markdown Artifacts: To ensure agents can consume content efficiently, markdown must be served via multiple methods: using a `.md` suffix, implementing content negotiation (via the Accept header), and utilizing a query parameter (`?mode=agent`).
- Web Content Pipeline (Web MCP): For advanced context retrieval, tooling should expose specific functions to agents, such as 'Search docs,' 'get pages,' and 'ask docs.'

## Practical implications

- For any documentation site or library with a public API, prioritize creating and maintaining agent-specific files (e.g., `LLMs.txt`, `AGENTS.md`) rather than relying solely on standard web content.
- Implement robust build steps to bundle markdown artifacts directly into the package's distribution directory (`node_modules`) to ensure agents have immediate access to context.
- When building a CMS, consider supporting multiple output formats (HTML, Markdown) and providing explicit instructions for agent consumption.

## Topics

Large Language Models (LLMs), Agent Architecture, Documentation Pipelines, Open Source Libraries, Web Development Best Practices, C15T, Inth, Aura AI

Source: https://www.youtube.com/watch?v=V_5bn4q-vAI