frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Vulnerabilities in Signal Sealed Sender and Usernames

https://eprint.iacr.org/2026/484
2•Arathorn•2m ago•0 comments

Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

https://writings.hongminhee.org/2026/03/legal-vs-legitimate/
3•dahlia•2m ago•0 comments

Current Large Audio Language Models largely transcribe rather than listen

https://arxiv.org/abs/2510.10444
1•PaulHoule•4m ago•0 comments

Show HN: Value stock screener built around Graham Number and Piotroski F-Score

https://stockpik.co
1•ecosystemj•4m ago•0 comments

Trajan: CI/CD Security Scanner

https://github.com/praetorian-inc/trajan
1•rdlbstr•6m ago•0 comments

Functional foods that lower bad cholesterol naturally (Evidence-based)

https://www.plantbasededit.com/12-functional-foods-to-lower-bad-cholesterol-naturally/
1•martinagabr•6m ago•0 comments

We are the data that woke up

https://twitter.com/atrupar/status/2030807244569252258
1•intentropy•6m ago•1 comments

The Arrogance of Ignorance. – By James Fallows

https://fallows.substack.com/p/the-arrogance-of-ignorance
1•MaysonL•6m ago•0 comments

88% of companies use AI. Only 13% trained anyone how

https://thoughts.jock.pl/p/ai-adoption-gap-who-actually-uses-ai-2026
2•joozio•7m ago•0 comments

Show HN: Nox – A tree-walking interpreted language written in pure Python

https://github.com/DevNexe/Nox
1•DevNexe•7m ago•0 comments

Why glibc is faster on some GitHub Actions Runners

https://codspeed.io/blog/why-glibc-faster-github-actions
6•art049•8m ago•0 comments

New farm bill would condemn pigs to a lifetime in gestation crates

https://twitter.com/Lewis_Bollard/status/2030985704902099335
21•bilsbie•9m ago•0 comments

Jetbrains: Air Launches as Public Preview – A New Wave of Dev Tooling

https://blog.jetbrains.com/air/2026/03/air-launches-as-public-preview-a-new-wave-of-dev-tooling-b...
1•virgildotcodes•10m ago•0 comments

The engine of Germany's wealth is blocking its future

https://europeancorrespondent.com/en/r/the-engine-of-germanys-wealth-is-blocking-its-future
2•mariuz•10m ago•0 comments

Frictionless by Nicole Forsgren and ABI Noda

https://www.sandordargo.com/blog/2026/03/07/frictionless
1•ibobev•11m ago•0 comments

We upgraded to a frontier model and our costs went down

https://www.mendral.com/blog/frontier-model-lower-costs
1•shad42•11m ago•0 comments

Show HN: NovusNet, an encrypted C++ networking library for beginners

https://github.com/Nullora/NovusNet
1•nullora•12m ago•0 comments

I'm Getting a Whiff of Iain Banks' Culture

https://probablydance.com/2026/03/07/im-getting-a-whiff-of-iain-banks-culture/
1•ibobev•12m ago•0 comments

A Primer on Bézier Curves

https://pomax.github.io/bezierinfo/
1•jihadjihad•12m ago•0 comments

Show HN: Free Coffee via Octopus Energy's Internal GraphQL API

https://frontbot.com/blog/how-i-automated-octopus-energy-coffee/
1•pi3rre•13m ago•0 comments

Perhaps not Boring Technology after all

https://simonwillison.net/2026/Mar/9/not-so-boring/
1•j4mie•13m ago•0 comments

Show HN: Wolf Defender, a open-weight prompt-injection detection model

https://huggingface.co/patronus-studio/wolf-defender-prompt-injection
1•patronusprotect•13m ago•0 comments

Show HN: Amux – single-file agent multiplexer for headless Claude Code sessions

https://amux.io
1•Beefin•13m ago•1 comments

Copilot Cowork: A new way of getting work done

https://www.microsoft.com/en-us/microsoft-365/blog/2026/03/09/copilot-cowork-a-new-way-of-getting...
6•daniel_iversen•14m ago•1 comments

The Burnout Economy

https://green.spacedino.net/the-burnout-economy/
2•surprisetalk•14m ago•0 comments

I Am Experimenting with Blocking HTTP1.1

https://sheep.horse/2026/2/i_am_experimenting_with_blocking_http1.1.html
1•surprisetalk•14m ago•0 comments

You're Either Excellent, Middling or Lost

https://aelerinya.substack.com/p/youre-either-excellent-middling-or
1•surprisetalk•14m ago•0 comments

Abundance of the Commons

https://supernuclear.substack.com/p/abundance-of-the-commons
1•surprisetalk•14m ago•0 comments

Show HN: TwoTicks – automated crypto trading strategies

https://twoticks.ai/
1•rebeccagaskell•15m ago•1 comments

Prefix sums at gigabytes per second with ARM NEON

https://lemire.me/blog/2026/03/08/prefix-sums-at-tens-of-gigabytes-per-second-with-arm-neon/
1•ibobev•18m ago•0 comments
Open in hackernews

Show HN: Agent Firewall – Go proxy to kill LLM death spirals

2•wuweiaxin•2h ago
I run 6 AI agents as my entire team. Yesterday two agents got stuck in an infinite loop arguing over JSON formatting. Burned $47 in API calls while I slept.

Anyone running multi-agent setups (CrewAI, AutoGen, LangGraph) knows the pain: agents go rogue, tokens burn, no circuit breaker.

Building an open-source Go reverse proxy. Change one env var (OPENAI_BASE_URL=http://localhost:8080/v1), it detects death spirals and physically cuts the connection.

Local-only, no SDK, no cloud, single binary. Ship in 2 weeks if enough people need it.

Have you been burned by runaway agents?

Comments

veloryn•2h ago
This kind of failure mode is going to become more common as people start wiring multiple agents together. Infinite loops, runaway API costs, agents triggering each other recursively etc. Most frameworks focus on capability but very little on runtime control. Curious to see if people start building proper guardrails or 'governance layers' around agents the same way we do with traditional distributed systems.
wuweiaxin•1h ago
Exactly. We're basically living through the "distributed systems reliability" moment for agents. Traditional infra has circuit breakers, rate limiters, health checks — agents have nothing.

That's what I'm building: treat agent-to-LLM calls like microservice calls. Detect loops, enforce budgets, fail-open if the firewall itself crashes. Same patterns, new domain.

The interesting part is loop detection needs to be semantic, not just frequency-based. Two agents can politely argue in circles without ever triggering a rate limit.