frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Low Poly Earth: A stylized low-poly Google Earth

https://play.lowpolyearth.online/
1•blueblazin•1m ago•1 comments

Immersive Cosmology

https://johnhoffman.io/posts/inside-the-cosmic-web
1•astro1234•2m ago•1 comments

Time-lapse: spider egg sac, 2012 [video]

https://www.youtube.com/watch?v=YMFmXwVsgbI
1•Eridanus2•2m ago•0 comments

A hacker's guide to bending the universe (2016)

https://aresluna.org/a-hackers-guide-to-bending-the-universe/
1•CharlesW•4m ago•0 comments

Pggo: Tiny, fast Postgres driver for agents

https://pggo.dev/
2•handfuloflight•4m ago•0 comments

5x faster Edge Functions: V8 isolates to Firecracker MicroVMs

https://www.netlify.com/blog/edge-functions-firecracker-microvms/
2•jbott•5m ago•0 comments

A year of databases: how I fell in love with programming again

https://dennypenta.github.io/mynameis/blog/year-databases-11
1•adletbalzhanov•6m ago•0 comments

Falcon Design

https://falcon.so/
2•handfuloflight•7m ago•0 comments

Show HN: Dbb1 radio – endless radio powered by Jev

https://radio.dbb1.dev/
1•dougbarrett•7m ago•0 comments

Fractional Is Just a Racket

1•artifix•7m ago•0 comments

Tesla patents 'Electric Fan Car' with 4 ducted fans ahead of Roadster reveal

https://electrek.co/2026/09/29/tesla-electric-fan-car-patent-model-s-roadster/
3•jethronethro•8m ago•2 comments

Book Review: Life and Death by Stephenie Meyer (2015)

https://booksatruestory.com/2015/10/20/book-review-life-death-stephenie-meyer/
1•Tomte•8m ago•0 comments

Museum of Time-Based Art

https://motba.art
1•NaOH•12m ago•0 comments

A glimpse into life inside a child prison in France

https://www.theguardian.com/global-development/2026/sep/28/i-sleep-with-all-my-clothes-on-to-prot...
2•rl3•14m ago•0 comments

Show HN: Make personalized cinematic videos in minutes

https://yume.video
1•dayaya•14m ago•0 comments

Another Look at Provable Security

https://www.math.uwaterloo.ca/~ajmeneze/anotherlook/
1•ibobev•15m ago•0 comments

The Elevator Receipt

https://www.distributedthoughts.org/2026-09-28-the-elevator-receipt/
1•speckx•15m ago•0 comments

Qt 6.12 LTS Released

https://www.qt.io/blog/qt-6.12-released
3•ibobev•15m ago•0 comments

DxTimingCaptureLibrary

https://devblogs.microsoft.com/directx/introducing-dxtimingcapturelibrary/
1•ibobev•16m ago•0 comments

Carnegie Mellon University Announces Historic $3B Gift from Ken Griffin

https://www.cmu.edu/news/stories/archives/2026/september/carnegie-mellon-university-announces-his...
2•Guyadou•16m ago•0 comments

Show HN: SideTap – an LLM drives a real iPhone from Windows, no Mac or jailbreak

https://github.com/ucsandman/SideTap
1•practicalsystem•17m ago•0 comments

Repack (Concurrently) in Postgres 19

https://boringsql.com/posts/repack-concurrently-costs/
1•plaur782•18m ago•0 comments

Reddit will stop supporting RSS feeds on November 13th

https://www.theverge.com/tech/1002788/old-reddit-ai-scraping
6•edsimpson•19m ago•1 comments

America.gov chatbot hallucinates Minecraft's End Poem

https://twitter.com/bennjordan/status/2105169596181590100
1•Aeroi•21m ago•1 comments

E2B Embed

https://github.com/e2b-dev/runtime/tree/main/embed
1•handfuloflight•21m ago•0 comments

Postgres {on,vs,with} Linux

https://www.youtube.com/watch?v=isATpooTax8
2•utiiiD•21m ago•0 comments

Show HN: We stress-tested LangGraph with an evolutionary fuzzer

https://github.com/zariffromlatif/life-forge
1•zariflatif•23m ago•0 comments

Interview with Richard Ho, OpenAI

https://morethanmoore.substack.com/p/interview-with-richard-ho-openai
1•ENOMEM•24m ago•0 comments

VideoArena – Generate Motion Graphics with various models and harness

https://videoarena.newmeasure.ai/
1•abhishek203r•24m ago•1 comments

F.02 Decommission

https://www.figure.ai/news/f-02-decommission
1•LiamPowell•25m ago•0 comments
Open in hackernews

Show HN: Token compression CLI to save Codex/Astra costs

7•yolandac•51m ago
Hey HN! Yolanda and Spencer here - wanted to share a token compression tool that we’ve built for ourselves to save 30% costs on codex!

After maxing out sub and burning $700/day per person on api, we fine tuned a compression model to trim codex's tool call output to reduce input token + cache. It cut down tokens by 29.6% and now I just leave it on by default in Codex.

To avoid messing up w/ cache, we use proxy + fine tuned qwen model trained on preserving agent trajectory to remove tool call results before they go back to the model, leaving kv cache untouched.

The cli is free for everyone to use (https://github.com/spenmcke/compress). Just lmk ur feedback and hacks to shave even more costs on astra! If you want to integrate it into your product to offer the best models at low cost, I can set you up with an sdk and api keys

PS: It’s built for coding agents, not conversational agents. I optimized it for file retrieval accuracy, trajectory preservation, and quality to get up to 30% cost reduction depending on how context-heavy the task is.

On security and privacy side, it's a proxy wrapping your local codex and ZDR so it doesn't retain any queries. It’s on by default in codex and when you don’t want compression, you can use `codex --uncompress` to disable it.

Give it a try: code is in https://github.com/spenmcke/compress

You can install the cli using

`curl -fsSL https://install.everestagi.com/install.sh | sh && source ~/.config/everest/shell.sh`

Love to hear any feedback and learn your hacky ways to save token costs too!

Comments

stdl1b•34m ago
how did you measure it's 30% saving?
yolandac•30m ago
we're a proxy so could count the tokens using openai response.usage
michaelastreiko•31m ago
Practical angle I like: a checkable token cut beats another flashy demo. For a small team, predictable spend on coding agents matters more than peak hype.
yolandac•26m ago
it gets addictive to run the `savings` command to see tokens saved