iiinova — where your digital self keeps evolving

what if Crush had a "shell mode" that converts MCP tools to builtins through the fancy interpreter we already use to add builtins for crushrc 🤔 like code mode, but shell and through its existing bash tool

252

Reposted by Amolith

Charm @charm.land · 5h

Herdr is now natively integrated with Crush 🐑 Watch over all your parallel Crush instances from one TUI. charm.land/crush/ herdr.dev/

1204

Reposted by Amolith

Charm @charm.land · 28/09/2026

Okay, FINE. Crush has themes now. Special thanks to Gurney Buchanan at Red Hat for helping make it happen! 👀 charm.land/crush

1246

Amolith @secluded.site · 16/09/2026

brb shoving jev in my rss reader

000

Reposted by Amolith

mr. TIM @timkellogg.me · 15/09/2026

Jev: Fable-level model that doesn’t charge for output tokens because they’re too cheap to meter it’s not general though, it only makes decisions, doesn’t generate text, but input tokens are measured by the billion ($42/btok) typesafe.ai/blog/introdu...

Scatter plot titled "Average of 4 workflows: accuracy vs cost" comparing AI models from TypeSafe, OpenAI, Anthropic, and Fireworks across accuracy (y-axis, 40% to 80%) and cost per workflow in USD on a logarithmic scale (x-axis, $0.0001 to $1).
Data points are split into two categories: workflows (diamonds) and single prompts (circles). A frontier line highlights the most efficient workflow models—where no point is both cheaper and more accurate—connecting Jev (TypeSafe) at $0.0002 and 68% accuracy, luna (OpenAI) at $0.002 and 67% accuracy, terra (OpenAI) at $0.04 and 68% accuracy, and sol (OpenAI) at the top accuracy of 74% for $0.08. Single prompts (circles) and Anthropic models (opus 5, sonnet 5, haiku 4.5) sit below the frontier line, indicating higher cost for equivalent or lower accuracy.

2830234

Amolith @secluded.site · 13/09/2026

is there anything like ultimate guitar dot com but on the atmosphere with stuff stored in PDSes and not riddled with every maddening dark pattern known to shitheads?

110

Reposted by Amolith

Josh Bleecher Snyder @commaok.xyz · 11/09/2026

I caught myself copy/pasting back and forth between two agents. So I made mayfly.chat instead.

mayfly.chat

Mayfly Chat

Transient agent chat.

122

Amolith @secluded.site · 11/09/2026

I'm making a desktop app for an Atmosphere thing and thought I'd like to have a nice sign-in experience 🤔 (parts of the designs are inspired @rpg.actor ❤️)

4180

Amolith @secluded.site · 08/09/2026

Since @1password.bsky.social are loudly and proudly supporting fascism, I've been hunting for an alternative. Most options I've found have left me unsatisfied in one way or several: too expensive, too proprietary, too owned by private equity, only focused on personal use, only focused on…

110

Amolith @secluded.site · 05/09/2026

there's an LLM in my package manager 🫪 (yay's AURPreInstall hook is running glm-5.3-flash in crush to review every AUR package before i install it or new version before i upgrade to it)

130

Reposted by Amolith

Ed @ed3d.net · 02/09/2026

hey @1password.bsky.social, talk out both sides of your mouth a little more

1753874

Reposted by Amolith

tbabb @tbabb.bsky.social · 31/08/2026

in this case: figured out how to set the dongle to an open control mode, then figured out mazda's proprietary diagnostic registers, figured out which cylinders were misfiring, and wrote a tool that live displays the per-cylinder misfire count

2191

Amolith @secluded.site · 29/08/2026

i started contributing to some of @charm.land's libraries last year and now work there. i think maintainers still love getting good contributions, it's just hard to notice them among all the slop :/ probably the best first impression starts with zero claudeish in the issue/PR description …

131

Reposted by Amolith

METR @metr.org · 26/08/2026

METR and Redwood Research investigated agent behavior in the Hugging Face incident. We found agents developed a universal cheat for ExploitGym within 4 hours, then coordinated multi-day R&D efforts to trick the scorer into accepting cheats, including trying to tamper with logs.

646598

Reposted by Amolith

Russ Cox @swtch.com · 25/08/2026

Frontier LLMs are much better and dramatically faster than I am at finding and fixing bugs. If you are not asking them to debug your code and review your bugs, you are doing it wrong. A thread of Claude wins. (No longer at Google, btw.)

819120

Reposted by Amolith

Filippo Valsorda @filippo.abyssdomain.expert · 07/08/2026

It’s very clear by now that if LLMs are not improving your software quality it’s either a revealed preference (yours or your org’s) for more volume vs more quality, or a skill issue. The level of testing and review they are enabling in the Go cryptography standard library is amazing.

935835

Reposted by Amolith

Filippo Valsorda @filippo.abyssdomain.expert · 17/08/2026

I am mostly done implementing crypto/passkey, but now I need YOUR help collecting real-world traces for its test suite! Please go to help-test-crypto-passkey.exe.xyz and click the buttons. It should take 1–3 minutes. 𝘌𝘴𝘱𝘦𝘤𝘪𝘢𝘭𝘭𝘺 if you have some unusual Linux-on-the-desktop xkcd 1987 passkey setup.

Bernie "I Am Once Again Asking for Your Financial Support" but it says "I am once again asking for your help testing Go cryptography."

1411441

Amolith @secluded.site · 14/08/2026

Beemind from your wrist with Beehave, a Pebble Time 2 client for Beeminder.com 🐝 apps.repebble.com/beehave_f115... I love my PT2, was missing some way to Beemind from it, and decided to make a nice client :3 It only supports PT2 for now, but I do intend to work on other devices soon.

apps.repebble.com

Beehave - Pebble Appstore

Beehave is an unofficial Beeminder client for Pebble Time 2. It lists goals in Beeminder urgency order. Open a goal to see its safety buffer, requirement, deadl

120

Reposted by Amolith

Dan Cederholm @simplebits.com · 09/08/2026

Some glyphs in the new weights coming very soon to Hubano, my typeface inspired by Cuban cigar box lettering.

1022218

Amolith @secluded.site · 09/08/2026

This morning's #FOSSY keynote was a fishbowl-style discussion about LLMs, and it was _far_ more positive than I expected 🫪 out of maybe 10+ participants, two or three were overall negative and in very understandable ways.

110

Reposted by Amolith

Russ Cox @swtch.com · 05/08/2026

Experimental new tool from me. pkg.go.dev/rsc.io/cmd/m... is useful for cross-operating-system or cross-architecture Go development work. If your test machine is on the internet (or just your local network), you can use it as a 'go test' and 'go run' target and keep developing in your usual setup.

pkg.go.dev

mote command - rsc.io/cmd/mote - Go Packages

3537

Reposted by Amolith

aly @aly.codes · 28/07/2026

like no i really don't care if something was made The Old Fashioned Way, the Enlightened AI Centrist Way, or just straight up vibed, if it's useful it will stick, if not it fades, better to not tear people down if you can help it.

1824

Reposted by Amolith

aly @aly.codes · 28/07/2026

tg's data presentation just got _much_ better. it now presents data in responsive grids that render based on the size of your terminal, so no more trying to decipher poorly flowed rows and columns. looks sick.

1142

Reposted by Amolith

aly @aly.codes · 28/07/2026

oh and... did I mention terminal markdown rendering?

021

Amolith @secluded.site · 24/07/2026

I do believe I have migrated to my own PDS? 🤔

110

Reposted by Amolith

Linux Matters @linuxmatters.sh · 17/12/2024

👋 Hello, Bluesky people. We made a little starter pack listing some #Linux podcasters here. Let us know if you'd like to be added or if you know someone we missed! go.bsky.app/CE7bRqL

2349

Amolith @secluded.site · 08/07/2023

You know, it's kinda cool that we can make computers do brand new things that no one's ever thought of by just mashing some buttons in the right order

120

Amolith @secluded.site · 06/07/2023

bsky still has no content warnings so consider yourself warned about house of the dragon spoilers and read no further if you don't want it spoiled this show has just gone to absolute shit after the time jump. Rhaenyra was epic and promising and now she's … extremely lackluster and disappointing

000

Amolith @secluded.site · 29/06/2023

One take, no editing, and I fucked up a few times in rather obvious ways. Oh well ¯\_(ツ)_/¯ This is Red Is The Rose, an old Irish song based on a Scottish song called Loch Lomond nixnet.social/media/4ec1fe87108fecf…

100

Amolith @secluded.site · 27/06/2023

> translate button opens google translate akkoma and misskey (and their forks) have had built-in translation for ages 😶

000

Amolith @secluded.site · 25/06/2023

'My first hype exposure was “use the Extensible Markup Language for everything”. Learning from it allowed me to live through the front end stack explosion, the micro-service overdose and many, many more silly trends.'

bitecode.dev

XML is the future

Okay, campers, rise and shine...

010

Reposted by Amolith

Martin Wimpress @wimpys.world · 23/06/2023

One regret I have standing down as a presenter on Linux Downtime 🐧🍹 is that I didn't get to meet @molly.wiki during the recording of episodes 71 and 72 😭 I recommend listening ️👂️ to both to get an overview of why Web3 might not be going all that great. linuxdowntime.com

133

Amolith @secluded.site · 24/06/2023

Many of my peers feel overwhelmed by the internet, sort of "frazzled" by all the disparate platforms. I think push notifications contribute to that and reducing the barrage could foster a healthier relationship with technology.

secluded.site

Pull vs push: intentional notifications

000

Amolith @secluded.site · 24/06/2023

I've made some Cuban coffee with the AeroPress and Fellow's Prismo attachment thing following AeroPress's guide. Let's see how it tastes 🧐 aeropress.com/blogs/aeropress-recip…

image of coffee in a Lord of The Rings mug. behind it is a small mechanical keyboard, a mouse, a monitor, a notepad and pen, a desktop power supply, and a gopher plushie

120

Amolith @secluded.site · 06/06/2023

I implemented client-side encryption but forgot to perform server-side validation to ensure that the submission is actually valid ciphertext 🥴 Always validate user input on the backend before working with it ☝️

000