Blog: Building the Most AI-Friendly Developer Resource

[0xYankee-Raccoons]

Summary

Long-form blog post covering everything we've done to make Jupiter's developer docs AI-friendly. The post is structured so each section can stand alone as a Twitter post for the devrel account.

Sections (each = 1 tweetable chunk)

1. Intro — Why AI-friendly docs matter (agents are the new developers)
2. llms.txt — Giving AI a map of your docs (dual-tier: index + full context)
3. Dual descriptions — Writing for humans AND machines (llmsDescription frontmatter)
4. MCP integration — Docs inside your AI editor
5. skill.md + Agent Skills — Machine-readable capabilities + open-source repo
6. REST-first, RPC-less — Making APIs agent-native by design
7. Multiple entry points — RAG, MCP, skills, markdown export — meeting AI where it is
8. What we learned — Principles for AI-friendly docs

Files

• blog/building-ai-friendly-docs.mdx — the blog post
• blog/index.mdx — add card for new post
• docs.json — add to navigation
• static/blog/building-ai-friendly-docs/ — image assets (TBD)

Publishing plan

• Post full blog to docs site + Twitter thread
• Every other day, repost the blog with main tweet highlighting one chunk