Hey HN! I built this because I was drowning in the AI agent ecosystem. Every week there are new frameworks, new MCP servers, new platforms — and the only way to find them was wading through "top 10" blog posts that are really just affiliate link farms.

So I built a single directory: 316 tools across 26 categories (frameworks, coding agents, MCP servers, enterprise platforms, DevOps agents, memory systems, evaluation tools, and more).

Each tool has: description, pricing model (free/freemium/paid/open-source), category tag, and direct link. You can search and filter by category, browse individual category pages, or read comparison articles on the blog.

Tech stack is deliberately simple: static HTML/CSS/JS with a tools.json data file, deployed on Vercel. No framework, no build step — just files. Category pages are generated with a Node script.

What's next: GitHub stars/popularity metrics, community voting, side-by-side comparison tables, RSS feed for new additions, and API access.

Some interesting patterns from cataloguing 316 tools: - Top categories: Platforms (36), Coding Agents (34), Dev Tools (29), Frameworks (25) - Underserved categories: Security (2 tools), Finance (1), Observability (1) - The biggest investment in AI agents right now is developer tooling, not chatbots

All feedback welcome — what tools am I missing? What categories should I add?