# AgentStackd > AgentStackd is an independent, hand-reviewed directory of 77 AI agents and automation tools, run by Wes Pierce. Each tool page lists the pricing model, pros, limitations and the vendor sources they were checked against. Listings are free. Paid options (fast-track review, category featured) only speed up review or add a clearly labeled slot; they do not buy a listing. Prices and plans change, so confirm on each vendor's site. ## MCP server (read-only) - [AgentStackd MCP endpoint](https://agentstackd.com/api/mcp): Remote MCP server (streamable HTTP, no auth). Tools: search_tools, get_tool, compare_tools, list_categories, new_releases. Every result includes the agentstackd.com page URL. - [How to connect](https://agentstackd.com/mcp/): Config for Claude Desktop, Cursor and VS Code. ## Buyer's guides - [The Best AI Coding Agents in 2026](https://agentstackd.com/best/ai-coding-agents-2026/): Nine coding agents grouped by where they run: your editor, your terminal, the cloud, or a prompt-to-app builder. - [The Best AI Research Tools](https://agentstackd.com/best/ai-research-tools/): Perplexity for cited web answers, Elicit and Consensus for papers, NotebookLM for your own documents. - [The Best AI Sales Agents & Tools](https://agentstackd.com/best/ai-sales-agents/): How Clay, Apollo, 11x and Lavender fit together, from enrichment to AI SDRs. ## Comparisons - [Cursor vs Claude Code (2026)](https://agentstackd.com/compare/cursor-vs-claude-code/): Editor or terminal? The real split is how much of the task you want to hand off. - [n8n vs Zapier (2026)](https://agentstackd.com/compare/n8n-vs-zapier/): Self-hosting and per-run pricing on one side, zero ops and the biggest app catalog on the other. - [Lovable vs Bolt.new (2026)](https://agentstackd.com/compare/lovable-vs-bolt/): Credits or tokens, which backend you get, and which one can import your existing repo. - [Cursor vs GitHub Copilot (2026)](https://agentstackd.com/compare/cursor-vs-github-copilot/): Move to an AI-first editor, or keep your IDE and let agents work from GitHub issues. - [ChatGPT vs Claude (2026)](https://agentstackd.com/compare/chatgpt-vs-claude/): Plans, bundled coding agents and research modes for the two most-used assistants. - [Zapier vs Make (2026)](https://agentstackd.com/compare/zapier-vs-make/): A step-by-step Zap editor against Make's visual canvas, and how each one meters usage. - [Claude Code vs OpenAI Codex (2026)](https://agentstackd.com/compare/claude-code-vs-openai-codex/): Two lab-made coding agents bundled into $20 chat plans. Surfaces, free options and API billing. - [Replit Agent vs Lovable (2026)](https://agentstackd.com/compare/replit-agent-vs-lovable/): Two prompt-to-app builders: plan prices, free options and how each one meters AI work. - [LangChain vs CrewAI (2026)](https://agentstackd.com/compare/langchain-vs-crewai/): Two MIT-licensed agent frameworks: components and LangGraph, or role-based crews and flows. - [All comparisons](https://agentstackd.com/compare/) ## Categories - [Coding Agents](https://agentstackd.com/category/coding/): Ship code with AI pair programmers and autonomous engineers. (20 tools) - [Research](https://agentstackd.com/category/research/): Find, read and synthesize sources faster. (9 tools) - [Sales](https://agentstackd.com/category/sales/): Prospect, enrich and write outreach on autopilot. (4 tools) - [Customer Support](https://agentstackd.com/category/support/): Resolve tickets with AI agents that actually close them. (6 tools) - [Content & Media](https://agentstackd.com/category/content/): Write, edit, voice and video, generated. (8 tools) - [Personal Assistants](https://agentstackd.com/category/assistants/): General-purpose AI assistants for everyday work. (15 tools) - [Automation & Agent Builders](https://agentstackd.com/category/automation/): Build your own agents and wire up workflows. (15 tools) ## Best for your job and AI stacks - [Best AI agents for your job](https://agentstackd.com/best/) - [AI stacks](https://agentstackd.com/stacks/) - [Top 3 picks per category](https://agentstackd.com/top-picks/) ## For makers - [Submit a tool (free)](https://agentstackd.com/submit/) - [Fast-track review](https://agentstackd.com/fast-track/): $39 for a decision in 48 hours, refunded if not accepted. - [Sponsor](https://agentstackd.com/sponsor/) - [Badge for makers](https://agentstackd.com/badge/) ## Optional - [About and how we pick tools](https://agentstackd.com/about/) - [Full tool list for LLMs](https://agentstackd.com/llms-full.txt) - [Sitemap](https://agentstackd.com/sitemap.xml)