frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Substack makes money from hosting Nazi newsletters

https://www.theguardian.com/media/2026/feb/07/revealed-how-substack-makes-money-from-hosting-nazi...
1•mindracer•35s ago•0 comments

A New Crypto Winter Is Here and Even the Biggest Bulls Aren't Certain Why

https://www.wsj.com/finance/currencies/a-new-crypto-winter-is-here-and-even-the-biggest-bulls-are...
1•thm•38s ago•0 comments

Moltbook was peak AI theater

https://www.technologyreview.com/2026/02/06/1132448/moltbook-was-peak-ai-theater/
1•Brajeshwar•1m ago•0 comments

Why Claude Cowork is a math problem Indian IT can't solve

https://restofworld.org/2026/indian-it-ai-stock-crash-claude-cowork/
1•Brajeshwar•1m ago•0 comments

Show HN: Built an space travel calculator with vanilla JavaScript v2

https://www.cosmicodometer.space/
1•captainnemo729•1m ago•0 comments

Why a 175-Year-Old Glassmaker Is Suddenly an AI Superstar

https://www.wsj.com/tech/corning-fiber-optics-ai-e045ba3b
1•Brajeshwar•1m ago•0 comments

Micro-Front Ends in 2026: Architecture Win or Enterprise Tax?

https://iocombats.com/blogs/micro-frontends-in-2026
1•ghazikhan205•4m ago•0 comments

Japanese rice is the most expensive in the world

https://www.cnn.com/2026/02/07/travel/this-is-the-worlds-most-expensive-rice-but-what-does-it-tas...
1•mooreds•4m ago•0 comments

These White-Collar Workers Actually Made the Switch to a Trade

https://www.wsj.com/lifestyle/careers/white-collar-mid-career-trades-caca4b5f
1•impish9208•4m ago•1 comments

The Wonder Drug That's Plaguing Sports

https://www.nytimes.com/2026/02/02/us/ostarine-olympics-doping.html
1•mooreds•4m ago•0 comments

Show HN: Which chef knife steels are good? Data from 540 Reddit tread

https://new.knife.day/blog/reddit-steel-sentiment-analysis
1•p-s-v•5m ago•0 comments

Federated Credential Management (FedCM)

https://ciamweekly.substack.com/p/federated-credential-management-fedcm
1•mooreds•5m ago•0 comments

Token-to-Credit Conversion: Avoiding Floating-Point Errors in AI Billing Systems

https://app.writtte.com/read/kZ8Kj6R
1•lasgawe•5m ago•1 comments

The Story of Heroku (2022)

https://leerob.com/heroku
1•tosh•5m ago•0 comments

Obey the Testing Goat

https://www.obeythetestinggoat.com/
1•mkl95•6m ago•0 comments

Claude Opus 4.6 extends LLM pareto frontier

https://michaelshi.me/pareto/
1•mikeshi42•7m ago•0 comments

Brute Force Colors (2022)

https://arnaud-carre.github.io/2022-12-30-amiga-ham/
1•erickhill•10m ago•0 comments

Google Translate apparently vulnerable to prompt injection

https://www.lesswrong.com/posts/tAh2keDNEEHMXvLvz/prompt-injection-in-google-translate-reveals-ba...
1•julkali•10m ago•0 comments

(Bsky thread) "This turns the maintainer into an unwitting vibe coder"

https://bsky.app/profile/fullmoon.id/post/3meadfaulhk2s
1•todsacerdoti•11m ago•0 comments

Software development is undergoing a Renaissance in front of our eyes

https://twitter.com/gdb/status/2019566641491963946
1•tosh•11m ago•0 comments

Can you beat ensloppification? I made a quiz for Wikipedia's Signs of AI Writing

https://tryward.app/aiquiz
1•bennydog224•12m ago•1 comments

Spec-Driven Design with Kiro: Lessons from Seddle

https://medium.com/@dustin_44710/spec-driven-design-with-kiro-lessons-from-seddle-9320ef18a61f
1•nslog•12m ago•0 comments

Agents need good developer experience too

https://modal.com/blog/agents-devex
1•birdculture•14m ago•0 comments

The Dark Factory

https://twitter.com/i/status/2020161285376082326
1•Ozzie_osman•14m ago•0 comments

Free data transfer out to internet when moving out of AWS (2024)

https://aws.amazon.com/blogs/aws/free-data-transfer-out-to-internet-when-moving-out-of-aws/
1•tosh•15m ago•0 comments

Interop 2025: A Year of Convergence

https://webkit.org/blog/17808/interop-2025-review/
1•alwillis•16m ago•0 comments

Prejudice Against Leprosy

https://text.npr.org/g-s1-108321
1•hi41•17m ago•0 comments

Slint: Cross Platform UI Library

https://slint.dev/
1•Palmik•21m ago•0 comments

AI and Education: Generative AI and the Future of Critical Thinking

https://www.youtube.com/watch?v=k7PvscqGD24
1•nyc111•21m ago•0 comments

Maple Mono: Smooth your coding flow

https://font.subf.dev/en/
1•signa11•22m ago•0 comments
Open in hackernews

Show HN: Plano – Edge and service proxy with orchestration for AI agents

https://github.com/katanemo/plano
8•adilhafeez•1mo ago
Hey HN — I’m Adil from Katanemo (with Salman, Shuguang, and Meiyu)

We previously shared an early version of this project as ArchGW. Based on customer feedback, the scope expanded from “LLM routing and model access” into something broader: delivery infrastructure for agentic applications. We renamed it to Plano and reworked the architecture accordingly.

The problem

On-the-ground AI practitioners will tell you that calling an LLM is not the hard part. The really hard part is delivering agentic applications to production quickly and reliably, then iterating without rewriting system code every time. In practice, teams keep rebuilding the same concerns that sit outside any single agent’s core logic:

This includes model agility — the ability to pull from a large set of LLMs and swap providers without refactoring prompts or streaming handlers. They need to learn from production by collecting signals and traces that tell them what to fix. They need consistent policy enforcement for moderation and jailbreak protection, rather than sprinkling hooks across codebases. And they need multi-agent patterns like handoff and specialization without turning their app into orchestration glue.

These concerns get rebuilt and maintained inside fast-changing frameworks and application code, coupling product logic to infrastructure decisions. It’s brittle, and pulls teams away from core product work into plumbing they shouldn’t have to own.

What Plano does

Plano moves core delivery concerns out of process into a modular proxy and dataplane designed for agents. It supports inbound listeners (agent orchestration, safety and moderation hooks), outbound listeners (hosted or API-based LLM routing), or both together.

Plano provides the following capabilities via a unified, protocol-native, framework-friendly dataplane:

- Orchestration: Low-latency routing and handoff between agents. Add or change agents without modifying app code, and evolve strategies centrally instead of duplicating logic across services.

- Guardrails & Memory Hooks: Apply jailbreak protection, content policies, and context workflows (rewriting, retrieval, redaction) once via filter chains. This centralizes governance and ensures consistent behavior across your stack.

- Model Agility: Route by model name, semantic alias, or preference-based policies. Swap or add models without refactoring prompts, tool calls, or streaming handlers.

- Agentic Signals™: Zero-code capture of behavior signals, traces, and metrics across every agent, surfacing traces, token usage, and learning signals in one place.

The goal is to keep application code focused on product logic while Plano owns delivery mechanics.

More on Architecture

Plano has two main parts:

Envoy-based data plane. Uses Envoy’s HTTP connection management to talk to model APIs, services, and tool backends. We didn’t build a separate model server—Envoy already handles streaming, retries, timeouts, and connection pooling. Some of us are core Envoy contributors at Katanemo.

Brightstaff, a lightweight controller written in Rust. It inspects prompts and conversation state, decides which upstreams to call and in what order, and coordinates routing and fallback. It uses small LLMs (1–4B parameters) trained for constrained routing and orchestration. These models do not generate responses and fall back to static policies on failure. The models are open sourced here: https://huggingface.co/katanemo

Plano runs alongside your app servers (cloud, on-prem, or local dev), doesn’t require a GPU, and leaves GPUs where your models are hosted.

Repo https://github.com/katanemo/plano + docs https://docs.planoai.dev/

Comments

dang•1mo ago
> We previously shared an early version of this project as ArchGW

Looks like these are the previous threads, if anyone's curious:

Show HN: ArchGW – An intelligent edge and service proxy for agents - https://news.ycombinator.com/item?id=44546265 - July 2025 (15 comments)

Show HN: ArchGW – An open-source intelligent proxy server for prompts - https://news.ycombinator.com/item?id=43259862 - March 2025 (7 comments)

Show HN: archgw: open-source, intelligent proxy for AI agents, built on Envoy - https://news.ycombinator.com/item?id=42187132 - Nov 2024 (14 comments)

paidev•4w ago
Cool project—Envoy dataplane + tiny routing LLMs is a solid combo for agent handoffs without the usual orchestration bloat.

We've wrestled with similar delivery pains building MCP tools for agents (Claude/ChatGPT/Cursor all love 'em now). Proxies like yours shine for LLM routing, but tool backends often drag with auth/setup. Your MCP integration in filters caught my eye—@leanmcp/auth decorator drops proper JWT validation (Cognito/Auth0/etc.) to 20 lines vs 600+ raw, auto-injects user context everywhere. (Disclosure: co-founder on LeanMCP.)

How's the Brightstaff fallback to static policies holding up in prod? Happy to chat agent war stories.