frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Anyone else like llms.txt for architecture docs?

1•jbdamask•1h ago
I'm a fan of using the llms.txt pattern of progressive disclosure for internal app documentation. Agents just produce more consistent code with these.

Two things I regularly use it for are documenting decision records, under app/docs/adr, and to manage AWS learnings (https://github.com/jbdamask/aws-learnings-library).

For example, in the case of ADRs, as the codebase grows bash tools aren't enough for the agent to get the big picture. If a new feature requires async processing, the agent can orient itself by checking the adr llms.txt, first, then designing according to project standards, e.g.

llms.txt - [0002 Distributed SQS+Lambda Worker Pipeline Over Single-Container Agent]: [Accepted] Rejected porting the local single-session agent pattern to AWS; Lambda time limits, the propose-review gate, and idle-cost requirements each independently require the distributed worker model.

0002-distributed-worker-pipeline-over-single-agent-lambda.md "Each ingest job is a message on an SQS queue. A Lambda pulls one message, runs the full normalize→file→propose pipeline, and writes the proposal to DynamoDB + S3. SQS provides at-least-once delivery and DLQ-after-N-receives for durability. Separate worker Lambdas handle merge, embed, maintenance, and research."

Before I started doing this, there would be design entropy for similar features where one agent would decide to throw everything in a single lambda, another would bake a container, and yet another would use SQS.

Do others use this pattern? Are there things people don't like about it?

Personalized gene therapy relieves epilepsy in two boys

https://www.nature.com/articles/d41586-026-02267-0
1•Brajeshwar•43s ago•0 comments

How to Manage Your Robot: Working Effectively with Coding Agents

https://johnjeffers.com/how-to-manage-your-robot/
1•johnj-hn•1m ago•0 comments

Codex swept my whole disk for credentials. grith froze every real one

https://grith.ai/blog/codex-credential-sweep-syscall-trace
3•edf13•4m ago•0 comments

Baker's Club: A short story about baking and community. Definitely not about AI

https://defnull.de/2026/bakers-club/
1•defnull•4m ago•1 comments

Can I Remix Madonna? Is This All Legal? An A.I. Music FAQ

https://www.nytimes.com/2026/07/24/arts/music/ai-music-faq.html
1•mitchbob•4m ago•1 comments

HN: SQL Guard – Stops Delete/Drop/Truncate Without Where in SSMS

https://sqlguard.app
1•manuelenzo•4m ago•0 comments

Values of College Training for Women

https://mtholyoke.com/dalbino/books/text/harpers01.html
1•geneticdrifts•6m ago•0 comments

Show HN: Power Coding – an AI-coding skill I audited on Claude Code and Codex

https://github.com/APareek89/power-coding
1•anandp1989•6m ago•0 comments

Postmortem.io – organized, searchable archive of real postmortems

https://postmortem.io/
1•dmonroy•7m ago•0 comments

'Igor, it's alive ' publicity stunts by the AI industry are becoming tedious

https://www.telegraph.co.uk/news/2026/07/24/ai-industry-out-of-control-publicity-stunts-tech/
2•arto•7m ago•0 comments

LLMs can write themselves notes to get better at reasoning

https://arxiv.org/abs/2607.20372
1•MarcoDewey•8m ago•0 comments

Researchers used electric stoves to treat asthma. It worked

https://www.climatecoloredgoggles.com/p/electric-stoves-asthma
1•speckx•9m ago•0 comments

Click to Pray, Click to Leak: Pope's Official App Exposes 700k User Emails

https://bobdahacker.com/blog/click-to-pray
1•miniBill•9m ago•0 comments

Jensen Huang on why open models matter

https://xcancel.com/JensenHuang/status/2080643682408321103?s=42
2•r3trohack3r•11m ago•0 comments

Half of social-science studies fail replication test in years-long project

https://www.nature.com/articles/d41586-026-00955-5
2•nradov•11m ago•0 comments

Writing a Debugger from Scratch

https://www.timdbg.com/posts/writing-a-debugger-from-scratch-part-1/
1•ibobev•12m ago•0 comments

Concurrent Servers

https://eli.thegreenplace.net/2017/concurrent-servers-part-1-introduction
1•ibobev•12m ago•0 comments

An OpenAI Model Escaped Its Sandbox. Where Was the Observer?

https://substack.norabble.com/p/an-openai-model-escaped-its-sandbox
1•nedruod•12m ago•0 comments

WebGPU Unleashed: A Practical Tutorial

https://shi-yan.github.io/webgpuunleashed/
2•ibobev•13m ago•0 comments

Show HN: Self-hosted multi-tenant agents built on Claude Code and Codex

https://github.com/OtoDock/oto-dock/
2•dimitrismrtzs•17m ago•0 comments

Trading Is All About Reading Charts

https://cryptodatamarket.com/
2•newworldwebs•17m ago•0 comments

Show HN: Preview and edit Material Kit React (MUI) without a build step

https://blog.crossui.com/2026/07/preview-and-edit-material-kit-react-without-a-build
2•linb•18m ago•0 comments

The FDA Could Clear These Peptides. Here's What the Research on Them Shows

https://www.nytimes.com/2026/07/23/well/peptides-fda-research.html
2•elo2000•18m ago•0 comments

Glm5.2.jsonl · HuggingFace/forensic-refusal at main

https://huggingface.co/datasets/huggingface/forensic-refusal/blob/main/glm5.2.jsonl
4•speckx•18m ago•0 comments

Ask HN: How much software do you build for yourself these days?

2•lellow•19m ago•1 comments

Open Weights and American AI Leadership [pdf]

https://images.nvidia.com/pdf/Open-Weights-and-American-AI-Leadership.pdf
5•lairv•20m ago•0 comments

Every market cycle tells a story

https://cryptodatacenter.io/
2•newworldwebs•20m ago•0 comments

Plants vs. Zombies for PlayStation 2

https://github.com/OptiJuegos/pvz-ps2
2•Jotalea•20m ago•0 comments

I checked whether "using AI" pays at YC startups. It doesn't

https://coderscreen.com/blog/does-using-ai-pay-at-yc-startups
2•rogutkuba•21m ago•0 comments

India Orders GitHub to Take Down Jack Dorsey's Bitchat Amid Protests

https://decrypt.co/374239/india-orders-github-to-take-down-jack-dorseys-bitchat-amid-protests
5•serendipty01•22m ago•1 comments