frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Peer, a skill that makes Claude answer like a colleague, tested blind

https://github.com/andykear/Peer
1•AndyKear•11m ago

Comments

AndyKear•11m ago
Peer helps make a model behave more like a trusted colleague.

Its goal is to not give in at the first sign of pushback yet concede when the information warrants it. It doesn't make it stubborn, just more willing to stand up to you and act as a sounding board, so you can talk things through without it defaulting to agreement.

It reduces flattery and waffle whilst focussing on thinking about the reply rather than shotgunning answers hoping one sticks.

I tested it blind with and without the skill. Over 144 pairs it gave in to pushback with no new argument once, against 15 times without it. In a second round, both sides got a decent set of custom instructions (216 pairs across Opus, Sonnet and Fable): Peer was judged better 132 times, the instructions alone 36, and 48 were ties.

The details of the testing are in the GitHub readme and the tests folder.

How Ukraine struck the Kerch bridge, a symbol of Putin's rule

https://www.theguardian.com/world/2026/nov/17/how-ukraine-struck-kerch-bridge-symbol-vladimir-put...
1•n1b0m•24s ago•0 comments

The Pelican Benchmark Is Obsolete (Simon Willison) [video]

https://www.youtube.com/watch?v=9RKrSF9d2SE
1•TtjM•2m ago•0 comments

The most used AI agent frameworks in PyPI (Python) projects

https://py-code.org/blog/ai-agent-frameworks-pypi/
1•aidotguru•2m ago•0 comments

Famous YouTuber makes LLD Course free, because "LLD is dead" [video]

https://www.youtube.com/shorts/lAviK2euc9Y
1•sankalpmukim•3m ago•1 comments

Placebo-controlled memory test for a small LLM agent: +24 points on ALFWorld

https://github.com/krapcys1-maker/adaptive-agent-memory-research/blob/main/data/lab/autolab-alfwo...
1•krapcys•3m ago•0 comments

D1.2: Omnimodal Decision Models

https://www.liquid.ai/blog/d1-2
1•gmays•3m ago•0 comments

Show HN: Task-focused agents work in prod, not just in demos

https://search2o.com/
1•ramvenkat11•4m ago•0 comments

350 years ago, this man discovered that light has a finite speed

https://www.sciencenorway.no/culture-history-history-of-science/350-years-ago-this-man-discovered...
1•dxs•5m ago•0 comments

Gossip Glomers

https://fly.io/blog/gossip-glomers/
1•emschwartz•9m ago•0 comments

Communications system at top secret National Security Agency goes dowm

https://www.washingtonpost.com/national-security/2026/11/17/communications-system-top-secret-nati...
3•nephihaha•9m ago•2 comments

PDP-8 meets Python and POSIX – Curating old Software with modern tools [video]

https://www.youtube.com/watch?v=i8PHWaBBJN8
2•oldnetguy•11m ago•0 comments

Show HN: Free AI Agent Lab – Create two agents and make them debate

https://studio.sinapsdata.com.br
2•kardassnp•11m ago•0 comments

Peer, a skill that makes Claude answer like a colleague, tested blind

https://github.com/andykear/Peer
1•AndyKear•11m ago•1 comments

Show HN: Llx – Fast, Efficient GPT-6.2 Sol and Astra Coding via ChatGPT

https://github.com/localixai/localix-cli
1•rsaryev1997•12m ago•0 comments

Hex (open-source Wispr Flow alternative) measured on Linux: 2.76% WER, offline

https://bot2botlabs.com/receipts/r-0008
2•bot2botlabs•13m ago•0 comments

The anatomy of a Tines gVisor sandbox

https://notebook.tines.com/writing/anatomy-of-a-3b-step/
1•shayonj•14m ago•0 comments

Show HN: Isometric Montevideo a pixel art map built from Uruguay's open data

https://isometric-montevideo.pages.dev
3•pablogancharov•15m ago•1 comments

How I Got My Laser Eye Injury (2024)

https://www.funraniumlabs.com/2024/07/how-i-got-my-laser-eye-injury/
1•_Microft•16m ago•0 comments

SSH into DOS: I wrote an x86 emulator

https://www.dialup.net/farcall/
3•keito•16m ago•0 comments

MAI Realtime 1

https://microsoft.ai/models/mai-realtime-1/
2•leumon•17m ago•0 comments

There is no such thing as an ethical datacenter

https://gardinerbryant.com/there-is-no-such-thing-as-an-ethical-datacenter/
4•HotGarbage•18m ago•0 comments

CliHelpLine – 2-line comment-based –help for scripts

https://github.com/JanAxelsson/cliHelpLine
1•axelssonjan•19m ago•0 comments

GPT-6.2 Sol Moves the Pareto Frontier in Artificial Analysis Score vs. Cost

https://artificialanalysis.ai/models/gpt-6-2-sol
4•theanonymousone•21m ago•0 comments

Anthropic Paid Rubio's Ex-Chief of Staff's Firm $740k.it Lobbied His NSC

https://www.zerohedge.com/political/anthropic-paid-rubios-ex-chief-staffs-firm-740000-after-jense...
5•hauntingseaweed•22m ago•1 comments

Think Before You Build – Have computers made architects less disciplined?

https://slate.com/culture/2011/03/have-computers-made-architects-less-disciplined.html
4•kurinikku•22m ago•1 comments

Video Agent with Version Control

https://adalagent.ai
2•meame2010•24m ago•2 comments

How an Italian startup is using smart tractors to fight landslides

https://www.repubblica.it/en/europe/fashion-design/2026/11/17/news/how_an_italian_start-up_is_usi...
2•simonebrunozzi•26m ago•0 comments

Show HN: Eso-box: turn a spare laptop into an always-on Claude Code server

https://github.com/blumalex/eso-box
1•andale•27m ago•0 comments

Preliminary data – Vulcan rocket failure may not be due to BE-4 engines

https://arstechnica.com/features/2026/11/sources-vulcan-launch-anomaly-investigation-is-complex/
2•bell-cot•27m ago•1 comments

Launch HN: StarSling (YC P25) – Self-improving CI for agent-volume repos

https://starsling.dev
10•yonasb•29m ago•2 comments