We got #1 Product of the Day on Product Hunt! 🏆
Huge thanks to everyone who supported us and believed in the future of AI agents. AgentKey will continue to help AI agents discover and use trusted tools, live data, and external capabilities.
The journey is just getting started.
Proud to be part of this next chapter at Chainbase.
A model can reason through a task, but completing it often means reaching beyond the model.
At AgentKey, we help AI access the tools and services it needs to get the job done.
The internet was built for people. Its next major users will be AI.
Today, we're introducing a new chapter for Chainbase.
From our roots in crypto data infrastructure, we're expanding our mission to build the Internet for Superintelligence, connecting AI to the information, tools, and services it needs to get things done.
Read our vision by Founder & CEO @moguxyz:
blog.chainbase.com/building-…
Open datasets become more useful when agents can work with them.
More teams can check results, ask new questions, and build on existing research without collecting the same data from scratch.
We'll explore bringing these datasets to @agentkeyapp.
Accurate weather forecasting starts with how clearly we can observe the atmosphere in real-time. 🌍
That’s why we’re releasing PARDIG and SatWeave - the foundational datasets behind WeatherNext 3 - to help researchers and developers build new climate applications 🧵
3/ Have a similar research question?
Connect your AI agent to AgentKey to look up competing products and compare contents, prices, coupons and delivery terms with source links you can check.
Start with 10 free research credits:
agentkey.app/
2/ Our October watchlist showed different ways to package value: e.l.f.'s $15 lip routine, themed collaborations, and Seelove's three-set offer for $64.99.
Comparing contents and bundle options gives a business more to test than matching the lowest price.
1/ A client was preparing for Black Friday and needed to understand competing makeup gift sets: what's included, how they're priced, and which offers are worth testing.
We've published the research and comparison tables here:
agentkey.app/blog/five-tacti…
6/ Tool discovery now better recognizes Chinese platform names and aliases. You can control capabilities by category, and execution results show the actual credits charged.
5/ Building with AgentKey? The Scenario API Beta offers 184 methods across 10 categories as of Oct 26.
Comparable tasks use standardized inputs and results. Methods with multiple sources can try a fallback without changing the method price.
4/ Company research now has more sources to work with.
RootData adds Web3 project and funding data. SEC / SEC-API, World Bank, and Eurostat expand financial coverage.
We've also expanded access to X, Xiaohongshu, WeChat content, Telegram, and automotive platforms.
3/ More sources for everyday questions:
Papers: alphaXiv, AMiner
Jobs: Liepin, JSearch
Travel: Booking, Rollinggo
Weather: QWeather, Google Weather
Agents can look up papers, compare job listings, or gather hotel, flight, and weather information for a trip.
2/ Use AgentKey in more of your existing tools.
New plugins and extensions for Codex, Kimi Code, Gemini CLI, Cursor, and Antigravity, plus MCP support for DeepSeek Harness.
Grok, Grok Bot, and Grok Build are also on our supported client list.
1/ AgentKey product update: July–October 2026
More supported clients, new data sources, and a Scenario API now in Beta.
Here's what we've added over the past few months.
Full update:
agentkey.app/blog/agentkey-p…
First MCP, now Agent Skills. Love to see it! 🔥
Open standards make it easier for developers to bring tools and workflows to more agent platforms, without starting over each time.
At Chainbase, we're working toward that through our data infrastructure and @agentkeyapp, helping agents access structured data, live information, and external tools.
The web should work for agents, too.
We're donating Agent Skills to the Agentic AI Foundation (@AgenticAIFdn), the same foundation we donated MCP to.
Since December 2025, 70+ agent platforms have integrated the standard, including ChatGPT, Cursor, GitHub Copilot, and VS Code.
Here's what that means for you:
A model with more tools is not automatically a better agent.
The real question is whether the agent can access the right tool, with the right context, at the right time.
That's why AgentKey focuses on a unified gateway for trusted tools and live data.
Harness Engineering could be the missing layer between smart models and reliable AI agents.
It gives agents context, safe tools, durable state, verification and recovery.
In this article, I show you how.
You’re probably using Claude wrong❌️
AI can be incredibly smart, but without access to the right tools and data, it’s limited.
@agentkeyapp changes that by giving your agent unified access to multiple tools.
Here’s how 👇
We make it easy.
A task often needs more than one data source. Your agent might search the web, check social activity, then look up onchain data.
With AgentKey, it can access all three without you setting up each service separately.
Robinhood's Agent Apps shows how data marketplaces can reach users through the agents they already use.
We're building in the same direction with AgentKey at Chainbase Labs: a marketplace of external data and tools, accessible through one connection.
That connection now covers 77 services and 1,000+ endpoints across search, social, crypto, finance, and more.
Over 20,000 users have signed up so far, and we're continuing to expand what their agents can access.