Blog: Engineering
We gave agents better instructions
We built a CLI command that gives coding agents current Mintlify guidance, and at first it changed nothing. One line in AGENTS.md, plus tighter guidance, took Claude Sonnet's pass rate on our hardest tasks from about one in five to every run.
We built one page to help AI agents set up Mintlify
We watched agents struggle through Mintlify setup, fetching multiple pages, trying wrong commands, and sending long instructions. So we built setup.md, one starting point for getting connected.
The Harness and Memory Architecture Behind Self-Updating Docs
How using fixed schema memory can improve agent performance at complex work.
Less is more: Improving editor performance
The Mintlify editor now opens, navigates, and edits very large pages up to 100x faster by doing less work.
Mintlify now supports WebMCP
Mintlify documentation sites now expose typed WebMCP tools that browser agents can use to search, read, and navigate documentation directly.
Scaling llms.txt
Large documentation sites were hitting Mintlify's 100,000-character llms.txt limit and omitting pages. We rebuilt generation as a hierarchy of files so agents can reach every page without loading the entire index.
Docs URL Benchmark: Markdown & llms.txt > HTML
We benchmarked four ways to serve documentation to AI agents (HTML, plain markdown, markdown linking to llms.txt, and markdown with llms.txt inlined) across 2,400 runs on 20 Mintlify docs sites, and found that a single link to llms.txt eliminates most agent 404s at no added cost.
How we built a virtual filesystem for our Assistant
We replaced expensive sandboxes with ChromaFs, a virtual filesystem over Chroma, to give our docs AI assistant the ability to explore documentation like a developer would.
Bridging two JSX runtimes: How we solved Astro's React children problem
How Mintlify solved a fundamental JSX runtime mismatch to make compound React components work inside Astro's MDX pipeline.
Auto-generating documentation sites from GitHub repos
How we built a multi-agent system to generate production-ready documentation sites from any public GitHub repository.
Declaring Clankruptcy: An experiment in agent orchestration
TL;DR: We tried agent orchestration, but an army of Claudes is still no match for one Dmytro.
Inside our effort to improve the Mintlify assistant
A data-driven look at improving the assistant, powered by ClickHouse and deeper feedback analysis.
How we eliminated cold starts for 72M monthly page views with edge caching
How we improved page speed performance across Mintlify documentation.
10 UI fixes I shipped in 10 days
Little details are what make UX so good you don't notice it.
What I shipped in my first 60 days at Mintlify
Building analytics infrastructure, abuse detection systems, and caching solutions.
Building an LSP for your docs
Adding an IDE experience to codeblocks in your docs is one of the issues that comes up often for people looking to add more context to code snippets. It's not a deal-breaker, but it can make your docs feel more connected to the functions and methods you're documenting.
We built our coding agent for Slack instead of the terminal
Behind the scenes of why we built our agent for Slack.
Debugging a mysterious HTTP streaming issue
We recently encountered a frustrating issue with HTTP response streaming at Mintlify. Our system uses the AI SDK with the Node stream API to forward streams, and suddenly things stopped working properly. The symptoms were confusing: streaming worked perfectly with cURL and Postman, but failed completely with node-fetch and browser fetch.
Refactoring mint.json into docs.json
Revamping a navigation schema might not sound flashy, but shipping docs.json is a major foundational update to mintlify.