BLUF Frontmatter Summaries
Every page includes a concise 1-2 sentence executive summary in its frontmatter description field. This description is automatically extracted for HTML meta tags, JSON-LD schemas, and the /llms.txt index.
Answer Engine Optimization (AEO) is the practice of structuring web content so AI-powered search engines (e.g., Perplexity, Gemini, ChatGPT, Claude) can easily read, parse, and cite it. This project implements AEO using a “Bottom Line Up Front” (BLUF) content structure, JSON-LD structured data (FAQ/About schemas), and automated LLM text indices (/llms.txt and /llms-full.txt).
Modern software engineers rarely browse websites page-by-page. Instead, they ask AI assistants questions like:
By optimizing this site for Answer Engines, we ensure that prospective candidates receive accurate, structured, and richly cited responses from their AI tools, drawing directly from our official sources of truth.
This project implements three primary layers of optimization:
BLUF Frontmatter Summaries
Every page includes a concise 1-2 sentence executive summary in its frontmatter description field. This description is automatically extracted for HTML meta tags, JSON-LD schemas, and the /llms.txt index.
JSON-LD Structured Data
The custom <Head.astro> component automatically compiles frontmatter metadata (description, techStack, successTraits, faqs) into standard schema.org JSON-LD FAQ/About scripts.
Automated LLM Text Indices
We serve two text-only endpoints: /llms.txt (a manifest of all pages with summaries) and /llms-full.txt (a single concatenated text archive of all content). These are automatically generated during builds (scripts/generate-llms.mjs) for RAG and LLM context windows.
When creating or editing pages, follow these guidelines to keep the site AI-friendly. Using our AI Skills (skills/team-page-builder.md & skills/blog-post-builder.md) handles all of this automatically!
Provide a 1-2 sentence executive summary in your page’s frontmatter description field:
description: "The Robotics System Team develops core motion planning algorithms and ROS2 controllers for Mujin's industrial robot arms."For team pages, ensure techStack and successTraits arrays are populated in your frontmatter:
techStack: - "C++20" - "Python" - "ROS2 / MoveIt"successTraits: - "Strong background in 3D kinematics" - "Experience with real-time systems"Do not edit public/llms.txt or public/llms-full.txt manually. The scripts/generate-llms.mjs script runs automatically during pnpm build, pnpm generate:llms, and git commit hooks to parse all MDX files and regenerate the LLM indices.