Agentry Tool Index: the AI agent tools in the news
In the 30 days to 21 November 2026, Agentry published 320 stories about AI agents; 55 of them name at least one of the 32 agent tools this index tracks. The most covered was Model Context Protocol (MCP) with 19 stories, followed by Gemini agents (10) and ChatGPT agents (8). AI crawlers and assistants fetched the stories about Model Context Protocol (MCP) 858 times.
Ranking
| Rank | Tool | Stories | 30 days before | Agent reads | Movement |
|---|---|---|---|---|---|
| 1 | Model Context Protocol (MCP) Anthropic / open standard | 19 | 31 | 858 | same |
| 2 | Gemini agents | 10 | 7 | 422 | up 1 |
| 3 | ChatGPT agents OpenAI | 8 | 10 | 340 | down 1 |
| 4 | Claude Code Anthropic | 6 | 3 | 264 | up 4 |
| 5 | GitHub Copilot GitHub / Microsoft | 6 | 5 | 214 | down 1 |
| 6 | Meta Muse Meta | 4 | 3 | 118 | up 3 |
| 7 | OpenAI Codex OpenAI | 3 | 0 | 124 | new |
| 8 | Moltbook Meta | 3 | 1 | 113 | up 8 |
| 9 | OpenClaw open source | 2 | 2 | 136 | up 4 |
| 10 | Cursor Anysphere | 2 | 1 | 84 | up 4 |
| 11 | LangChain and LangGraph LangChain Inc. | 2 | 1 | 82 | up 4 |
| 12 | OpenAI Dots OpenAI | 1 | 2 | 62 | same |
| 13 | ElevenLabs ElevenLabs | 1 | 0 | 33 | new |
| 14 | Grok xAI | 1 | 2 | 32 | down 4 |
| 15 | watsonx Orchestrate IBM | 1 | 4 | 31 | down 10 |
Latest stories per tool
Model Context Protocol (MCP)19 stories
Gemini agents10 stories
ChatGPT agents8 stories
Claude Code6 stories
GitHub Copilot6 stories
Meta Muse4 stories
OpenAI Codex3 stories
Moltbook3 stories
OpenClaw2 stories
Cursor2 stories
LangChain and LangGraph2 stories
OpenAI Dots1 story
ElevenLabs1 story
Grok1 story
watsonx Orchestrate1 story
Tracked, no stories in this period
Agentforce, Botpress, Browse AI, Character.AI, Chatbase, CodeRabbit, Devin, Lovable, Make, Manus, MindStudio, n8n, OpenAI Atlas, Perplexity Comet, Relevance AI, Runpod, Taskade
Method
The index counts the stories Agentry published in the last 30 days that name a tool in the headline or summary, and compares them with the 30 days before. Tools are ranked by number of stories, then by how often AI agents read those stories, then by name.
It measures coverage and agent interest, not quality. A tool ranks high because a lot happened around it, good or bad: launches, incidents, lawsuits and studies all count.
"Read by AI agents" counts fetches by known AI crawlers and assistants, MCP clients and readers of the markdown view. Human visits are not counted.
The list of tracked tools is fixed and maintained by hand; a tool that is not on the list is not counted. Sponsored articles are left out. Nothing on this page is paid for.
Data as JSON: /tool-index.json. Markdown version for AI clients: /agent/tool-index. Editorial standards: AI policy.
Cite as: Agentry. "Agentry Tool Index, 30 days to 21 November 2026." https://agentry.news/tool-index