Here is everything that happened in AI Agents this week 🧵 (save for...

“AI agent startup /dev/agents has raised a massive $56M seed round at a $500M valuation” - @TechCrunch
@dps alongside his cofounders @hbarra, @alcor, and @ficus. In addition to partnering with @ninaachadjian + @jillchase124

@braelyn_ai from @AgentOpsAI shared that “the amount of money people are willing to pay for agents im building in 5 minutes with AgentStack is astounding”

@hcompany_ai launched Runner H, the most advanced AI agent for real-world applications.
It doesn’t just outperform competitors in speed and accuracy—it also handles a far wider range of tasks, tackling challenges others can’t.

You can now clone any app just by pasting the link into Replit Agent.
- Paste the link
- Hover over it
- Click Take screenshot
Thanks for sharing @itsPaulAi

- Paste the link
- Hover over it
- Click Take screenshot
Real game-changer for building interfaces without coding.
@AlexReibman did a speedrun building an AI agent with AgentStack + @AgentOpsAI

Any% 2:52 [WR]
@NVIDIA_AI_PC introduces Agentic AI: Your ultimate taskmaster.
From managing schedules to automating workflows, Agentic AI is powered by RTX GPUs on local PCs with apps like @AnythingLLM for blazing speed and full privacy.

From managing schedules to automating workflows, Agentic AI is powered by RTX GPUs on your local PC with apps like @AnythingLLM for blazing speed and full privacy. 🔐
Check out this week's #AIDecoded blog: nvda.ws/3OsMJi1
Microsoft released a 100% open source multi-agent simulation framework!
Get instant feedback from 100 different AI personalities instead of spending weeks on user research.
Thanks for sharing @Saboo_Shubham_

Now you can get instant feedback from 100 different AI personalities instead of spending weeks on user research.
100% opensource.
@awscloud released a new Multi-Agent AI framework that lets you manage multiple AI agents, dynamically route LLM queries, maintain context across AI Agents and can be deployed locally on your computer.

It lets you manage multiple AI agents, dynamically route LLM queries, maintain context across AI Agents and can be deployed locally on your computer.
100% opensource.
@KaranVaidya6 launched AgentAuth by @composio - the most comprehensive auth solution designed for AI agents!
It supports 250+ apps across categories such as CRMs, ticketing, productivity, etc.

We understand the pain every developer experiences regarding authentication—managing OAuth flows for Gmail, handling API keys for Linear, or setting up permissions across multiple services. It's complex enough with traditional apps.
However, AI agents add an extra layer of complexity. Traditional auth solutions were not designed with this agent-specific requirement in mind.
Think about it—say you're building an agent to pull data from Salesforce, send updates through Slack, and log issues in GitHub. Each service requires different authentication—OAuth2, API keys, you name it—and your agent needs to work seamlessly on your users' behalf. Building and managing all that is a massive headache, right?
That’s precisely where AgentAuth by @composio comes in.
• It supports 250+ apps across categories such as CRMs, ticketing, productivity, etc.
• Compatible with 15+ Agentic Frameworks, including @langchain, @llama_index, @crewAIInc etc
• Offers self-hosting and white-labeling options.
• Provides a unified dashboard to monitor user accounts.
It takes care of all the complex authentication flows—OAuth, API Key, Basic, JWT, token refresh, and more—in the backend, so you can focus on building what truly matters for your users.
This is ideal for AI developers building real-world AI automation involving multiple application interactions.
Check out AgentAuth now - composio.dev/agentauth/
With the new @cursor_ai agent, you can now customize the agent w/ callable tools!
Thank you for sharing @NickADobos

It can run any python func in your codebase. The idea being being to check output & debug, but fuck that
CRAZY SIDE EFFECT:
this means you can customize the agent w/ callable tools! Like GPT Store actions or Claude MCP
Cursor Agent Builder!
Praxis AI (YC F24) builds agents capable of reasoning and processing complex situations in manufacturing. Their engineering copilot predicts and prevents machine failures in real time by solving problems the same way an engineer would.

ycombinator.com/launches/MKz-p…
@rom1trs shares an AI Agent that…
- exports Meta ads
- import into google sheet
- analyzes top performing ads
- iterates on hooks with ChatGPT
- generate winners iterations with arcads ai

> export Meta ads
> import into google sheet
> analyzes top performing ads
> iterates on hooks with chat gpt
> generate winners iterations with arcads ai
Our good friends at @AthenaIntell built Cursor for analysts using @copilotkit, @langchain & @tavilyai
“The Agentic-Copilot design pattern we've seen in apps like Cursor & Replit enables the best of human + AI collaboration.”
@BrendonGeils

🛠️Built using @copilotkit, @langchain & @tavilyai
The Agentic-Copilot design pattern we've seen in apps like Cursor & Replit enables the best of human + AI collaboration.
And it's coming everywhere - not just dev tools.
@dvdsgl built an AI Agent in just 5 minutes for evaluating candidates at @glideapps

Cursor Agent vs. Windsurf Agent
@omarsar0 has been testing both @cursor_ai and @windsurf_ai agents to build web apps.

I've been testing both Cursor and Windsurf agents to build AI web apps.
I believe that on the agent stuff, Windsurf is one step ahead.
The agent feature in Windsurf feels very native and like a first-class citizen.
My full demo and test here: youtu.be/6swSIMY6iTU
I noticed that Cursor's agent still struggles with very basic things like figuring out the right models to use for the AI web apps. It's also not very consistent.
I am not counting Cursor out. I am sure they can improve things really fast. I find it super interesting that even better and more powerful code editors are still on the horizon. Early days.
What has your experience been?
I share my perspective on “The biggest breakthrough in agents will be when they go from being reactive to proactively”
Actively hiring at @AgentOpsAI for Agent Engineers, AgentStack Engineers, and Evals Engineers. DMs are open.

I do not want to build workflows, I want an AI to tell me what workflows it can automate
@JiquanNgiam made an AI agent that helps with shopping by doing research about a person, suggests gifts, and then looks up good recommendations for them.
x.com/jiquanngiam/st…
“The big inflection point in building software comes when non techies can do technical things by using AI” - @bindureddy

Non-techies can already write complex SQL, create RAG chatbots and simple agents by simply speaking to AI in English
The next step will be more complex use cases
@hbjORbj plans to build an agent using AgentStack by @braelyn_ai + @AgentOpsAI. Benny went ahead and bought domain. Excited to see this progress.

“Langflow is the best tool I've found for building AI agents.” - @svpino
Things that make Langflow awesome:
1. Huge list of reusable components
2. Different templates to get things started
3. Agent-specific components and capabilities

Full stop.
If you are into agents, here are the things that make Langflow awesome:
1. Huge list of reusable components
2. Different templates to get things started
3. Agent-specific components and capabilities
1/3
@abacusai Introduces @github automations
- Using AI To write Code, Fix code
- Using AI to create Pull Request
- from the same Chat Interface (ChatLLM)
@MervinPraison

Using AI To write Code, Fix code
Using AI to create Pull Request
from the same Chat Interface (ChatLLM)
📚 Key Features
⚙️ How to Install?
🚀 How to Use?
💡 Key Benefits
🔄 Workflow Steps:
@abacusai
Control a computer with just your voice using @hume_ai
Built on @Replit's template to combine Hume’s empathic voice interface with @AnthropicAI's computer use API.
@sawyerhood taught a browser agent to watch youtube videos.
The next step is asking it to follow a tutorial.
x.com/sawyerhood/sta…
Did I miss anything? Share below.
Liked this thread? Share it with your friends/team.
Follow @adamsilverman and @AgentOpsAI for everything AI agent related.
Last weeks thread:
x.com/AtomSilverman/…








