Bookmarks

a short post

Behavioral differences from Claude Opus 5 and the prompting and harness patterns that address them: effort calibration, thinking behavior in API integrations and chat, progress updates, unattended and multiagent tasks, safeguard refusals, frontend design, complex visual inputs, multi-app workflows, and pasted text in user messages.

Build native apps in JS. Backend, frontend. ~5Mb. Contribute to tarwin/tinyjsapp development by creating an account on GitHub.

Build macOS, Windows, and Linux desktop apps in JavaScript: txiki.js backend, native webview window, ~6 MB shipped. No Electron. No ports.

Learn how OpenAI evolved Habitat from a Python library into a globally distributed storage platform serving 1 billion ChatGPT users and 22M requests per second.

One of the interesting challenges of the AI ecosystem in 2026 is that new, effective patterns emerge faster than I can adopt them. I’ll find a handful, get back to work, and realize a month later that I’d missed four or five more. The adoption cycle for Imprint this year has been something like: January: get every engineer onto Claude Code every single day March: ok, let’s also get everyone else onto Claude Code or Claude Cowork every single day April: local development is bottlenecked on checkout and worktree model, instead create ~10 local workspaces which each have an independent checkout of every repository, and operate at the workspace level, not at the repository level, so it can generate cross-repository pull requests across frontend, backend, infrastructure and data monorepos June: oh boy, agent-driven development is heavily constrained by lack of a common task management system with higher visibility and less permission complexity than Jira, so let’s migrate the entire company over to Linear and hard stop on Jira July: yikes, now we have visibility into all these tickets, many of them are trivial but managing them through local development isn’t scaling, let’s roll out an orchestrated harness which internally we call “Agent Fleet”, along the lines of Stripe’s Minions The most recent question for me has been figuring out how to adopt the software factory pattern. (After some light research, the specific AI-context origin of this term is slightly messy to attribute, but I think it might be Justin McCarthy in February 2026’s Software Factories And The Agentic Moment.)

We cut PR wait time and running costs by rethinking CI as a system, from the infrastructure underneath it to how work gets scheduled and tests get parallelized.

tanteki はその文書が本質的に述べたいことを見定めて、必要な情報だけを記述します。使用前後を実際の出力で見比べられます。

Contribute to iwasa-kosui/tanteki development by creating an account on GitHub.

Lovable previews are moving to OJ, a Rust-native dev server that starts them several times faster on a fraction of the memory. We are rolling it out gradually, and today OJ is open source in the Lovable GitHub organization.

Framework-neutral foundation for building generic DevTools. - devframes/devframe

Devframe is a framework-neutral foundation for devtools. One definition becomes a Web Standard handler you can mount into any host framework, ship as a CLI or static report, and expose to coding agents.

on proving yourself, burning out, and being a good teammate

tiny Jev-like family of decision models built on top of Qwen3.5 you can train and run on your own - jaredpalmer/kev

You can now use cache-mode to apply least-privilege access to the GitHub Actions cache at the workflow or job level. By granting each workflow or job only the cache access…

Last week TypeSafe AI unveiled Jev, their first example of a new category of model that they are calling “System One models” (I’m with Maggie Appleton, I think “decision models” …

MCP server portals in Access.

TIN is a fast, full-featured, full-text search index for Postgres

Declare an agentic task in YAML. AX sandboxes it, wires up its workspace, fences its network, and helps running it at scale.

Many important biological functions arise not from single genes, but from complex interactions encoded by entire genomes. Genome language models have emerged as a promising strategy for designing biological systems, but their ability to generate functional sequences at the scale of whole genomes has remained untested. Here, we report the first generative design of viable bacteriophage genomes. We leveraged frontier genome language models, Evo 1 and Evo 2, to generate whole-genome sequences with realistic genetic architectures and desirable host tropism, using the lytic phage ΦX174 as our design template. Experimental testing of AI-generated genomes yielded 16 viable phages with substantial evolutionary novelty. Cryo-electron microscopy revealed that one of the generated phages utilizes an evolutionarily distant DNA packaging protein within its capsid. Multiple phages demonstrate higher fitness than ΦX174 in growth competitions and in their lysis kinetics. A cocktail of the generated phages rapidly overcomes ΦX174-resistance in three E. coli strains, demonstrating the potential utility of our approach for designing phage therapies against rapidly evolving bacterial pathogens. This work provides a blueprint for the design of diverse synthetic bacteriophages and, more broadly, lays a foundation for the generative design of useful living systems at the genome scale. ### Competing Interest Statement B.L.H. acknowledges outside interest in Arpelos Biosciences and Genyro as a scientific co-founder. S.H.K. and B.L.H. are named on a provisional patent application applied for by Stanford University and Arc Institute related to this manuscript. All other authors declare no competing interests. Arc Research Institute, https://ror.org/00wra1b14 Stanford Institute for Human-Centered Artificial Intelligence

Valibot は vitest --typecheck をフル活用しているライブラリで、*.test-d.ts という型テスト専用のファイルが 250 ファイル・約 16,000 行 あります。ところで Valibot は @ts-expect-error を使い、以下のようなテストをしています。 test("sho…

Inside the industry's growing alarm—and the struggle to turn warnings into action.

How to prompt Opus 5.5, steer a long run, and check your results in Claude apps and Claude Code.

HarnessRouter Community Edition: the self-hosted, Apache-2.0 edition of the unified interface for agent harnesses. Run Codex, Claude Code, Hermes, PI, DSH, and more through one API, with sessions, ...

A reverse engineer's binary Swiss-army knife in Rust: triage, disassembly, function/CFG recovery, crypto-constant + YARA scanning for PE/ELF/Mach-O. Installs as 'knife'. - bl4ckr0ss3/knife

MCP サーバーが提供するツールを使いこなすには、複数のツールを組み合わせる手順や判断基準も必要です。Skills Extension は、こうした知識を記した Agent Skills を MCP 経由で発見・取得するための拡張です。この記事では設計の背景、スキルを段階的に読み込む仕組み、Node.js での実装例と対応状況を紹介します。

What Active Storage does at scale: variants built inside the request, the preprocessed trap, AVIF measured 31.6% under WebP, and the CDN check we failed.

Beskar 0.2.0 brings database-coordinated authentication limits, durable session revocation, and auditable administration to Devise and Rails built-in authentication.

Semantic decisions as ordinary Ruby #feels?, #decide, #score, Rails validations and pattern matching powered by Jev - Qew7/jev-feels