// the find
thedaviddias/llms-txt-hub
🤖 The largest directory for AI-ready documentation and tools implementing the proposed llms.txt standard
A directory site for websites implementing the llms.txt proposed standard — a convention for exposing structured documentation to LLMs, similar in concept to robots.txt. It pairs a browsable Next.js web app with a CLI tool that installs llms.txt docs as context into AI coding agents. The audience is anyone who wants to either publish their own llms.txt or discover what docs are available for the tools they use.
The CLI is the genuinely useful piece here — `llmstxt-cli` installing docs-as-skills into 35+ AI coding agents solves a real daily workflow problem. The monorepo is well-organized: Biome for linting, Changesets for versioning, Playwright for e2e, and there's actual security surface covered (CSRF, rate limiting, input sanitization all have dedicated test files). The GitHub Actions automation — workflows for updating the llms list, running link checks, and auto-merging approved submissions — means the directory stays current without manual intervention. Having the companion Chrome extension, VS Code extension, and MCP explorer gives the project multiple useful entry points beyond the website.
The directory has a curation problem that's getting worse: the README lists a Serbian locksmith, a Miami dry cleaning pickup service, an essay-humanization tool designed to evade Turnitin, and an HVAC company alongside Anthropic and ElevenLabs. The submission workflow is PR-based with bot automation, and quality control is clearly not keeping up. The whole thing is predicated on llms.txt becoming a meaningful standard, but right now it's a proposal — if it doesn't get adopted by tooling that actually reads these files in a useful way, the directory is a list of mostly-empty text files. The Next.js app with Supabase auth, favorites, user profiles, and community features seems significantly over-engineered for what is essentially a searchable link directory; that complexity is a maintenance burden that will compound as the team is presumably small.