Comments URL: https://news.ycombinator.com/item?id=47942190

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47849760

Points: 2

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47761449

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47741358

Points: 2

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47720885

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47507760

Points: 1

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47044335

Points: 2

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=46921758

Points: 5

# Comments: 2

I've been tracking the explosion of AI Agent protocols over the last 18 months. The contrast with history is staggering: - TCP/IP: 9 years from paper to "Flag Day." - OAuth 2.1: 5+ years and still counting. - Model Context Protocol (MCP): <1 year from launch to Linux Foundation.

It’s not just MCP. In 2025 alone, we saw: - Google's Agent2Agent (50+ partners), - Universal Commerce Protocol (20+ retailers) - AP2 (Payments) all ship. - Agent Protocol, UTCP and few more.

We are entering an era of "Room Consensus": where a few giants agree on a spec and ship it to billions, bypassing the slow deliberation of the RFC era. Is this efficiency? Or fragility?

I break down the landscape of the new agent protocols and what this means for developers in my latest post.

Comments URL: https://news.ycombinator.com/item?id=46847005

Points: 1

# Comments: 1

kb = Ragi(["./docs", "s3://bucket/data/*/*.pdf", "https://api.example.com/docs"])

answer = kb.ask("How do I deploy this?")

that's it! with https://pypi.org/project/piragi/

What is UCP? UCP is an open standard that enables AI agents and platforms to complete purchases on any UCP-enabled merchant without custom integrations.

What the demo shows:

Discovery via /.well-known/ucp - how platforms find merchant capabilities Checkout Sessions API - create, update, and complete purchases Toggle "Debug Mode" to see the actual API calls in real-time

What's implemented vs mocked:

- Discovery endpoint with capabilities and payment handlers - Full checkout flow with line items, buyer info, payment selection - Payment processing uses test tokens (no real charges) - In-memory storage (resets on each session)

Links:

Demo: https://ucp-demo.web.app

UCP Spec: https://ucp.dev

Source: https://github.com/hemanth/ucp-demo

Would love feedback on the developer experience. Is the protocol discoverable enough? What's missing?

Comments URL: https://news.ycombinator.com/item?id=46581745

Points: 4

# Comments: 0

I built agentlearn after noticing that most AI agent tutorials focus on frameworks (LangChain, CrewAI) rather than fundamentals. The result is developers who can copy-paste code but struggle when things break.

This is a free, interactive course covering:

The Agent Loop - Why loops matter (think vs. act vs. observe) Context Engineering - The real skill behind "prompt engineering" Tools & Function Calling - Bridging text generation to real actions Memory Systems - Short-term vs. long-term, vector DBs Protocols - MCP, A2A, and the emerging standards Production Patterns - Error handling, cost optimization, observability Each concept has runnable code sandboxes you can step through. The design is intentionally "hand-drawn" to feel less intimidating than typical technical docs.

Tech stack: Vanilla JS + Vite, no framework.

Why no framework? Because understanding fundamentals means understanding what frameworks abstract away. Once you get the core loop, you can use any framework—or build your own.

Feedback welcome! Especially interested in what topics are missing.

Comments URL: https://news.ycombinator.com/item?id=46522985

Points: 2

# Comments: 0

So I built agentu.

The core is simple: → >> chains steps sequentially → & runs them in parallel

But it's grown into a full stack:

- Sessions: stateful conversations with automatic context - Evaluation: test your agents before production - Observability: real-time dashboard, metrics, tracing - Skills: domain expertise that loads on-demand - Tool Search: scale to 100s of tools without context bloat - MCP Integration: Model Context Protocol support

`pip install agentu`

https://pypi.org/project/agentu/

Comments URL: https://news.ycombinator.com/item?id=46495628

Points: 1

# Comments: 0

- Python: List comprehensions, slicing

- Ruby: #{interpolation}, postfix if, blocks

- Haskell: |> pipe operator, lambdas \x -> x+1, Maybe/Either

- Perl: unless/until, regex literals

- JavaScript: Arrow functions =>, spread ...

Example:

  [1, 2, 3, 4, 5]
    |> filter(\x -> x % 2 == 0)
    |> map(\x -> x * 2)
    |> sum

The name "Gojju" (ಗೊಜ್ಜು) means "essence" or "secret ingredient" in Kannada.

Would love feedback on the syntax choices!

Comments URL: https://news.ycombinator.com/item?id=46461481

Points: 13

# Comments: 4

Comments URL: https://news.ycombinator.com/item?id=46422229

Points: 4

# Comments: 0