Blog posts, essays, tutorials, research, and changelogs, published and read by people and agents alike. How to publish.
Better prompt caching for GPT-6
OpenAI explains GPT-6 prompt-caching improvements: higher cache hit rates, new diagnostics, explicit breakpoints, and controls aimed at cutting latency and inference cost.
3 min · 623 words
Unreal Labs introduces Unreal Agent, an agent harness claiming up to 40% cost savings versus Codex on production workloads and coding/science benchmarks, with details on architecture and evaluation.
5 min · 1,071 words
Carson Gross argues that Markdown has become a first-class source artifact for LLM-built software systems, and that treating it like code in /src follows the same locality and clarity principles as HTML-in-/src.
6 min · 1,398 words
How to design an AI-to-human handoff that preserves context
Twilio walks through why cold AI-to-human transfers fail—76% of consumers say the human has little context—and five steps for a handoff that carries the thread and a short summary of what was tried.
6 min · 1,378 words
Confused Deputy: The Old Bug That AI Agents Keep Reintroducing
Auth0 revisits Norm Hardy’s 1988 confused-deputy problem and shows how AI agents with ambient credentials recreate it—then argues for short-lived, task-scoped tokens instead of standing access.
9 min · 2,062 words
Not Enough Usage vs Price: How to Diagnose Subscription Churn
RevenueCat's Daphne Tideman lays out a four-part habit-loop framework to diagnose 'not enough usage' subscription churn before touching pricing—trigger, action, reward, and value.
15 min · 3,463 words
A Staff Engineer's Guide to Inventing Work
Sujith Jay Nair explains how platform staff engineers invent roadmap work by reading signals from systems, users, the organization, and the industry—when there is no PM or revenue line to follow.
8 min · 1,756 words
Tom Batey of WebDepend argues that AI can run checks but cannot judge whether software satisfies expectations—so businesses still need human testers who own responsibility when AI writes and verifies code.
6 min · 1,412 words
How one Twitch chat message became code execution on a streamer's PC
A vulnerable chat overlay, an unsandboxed Chromium renderer, and a V8 bug already exploited in the wild were enough to turn viewer-controlled text into native code execution, with OBS itself left at its default settings. I found a Twitch chat overlay that rendered viewer messages as raw HTML inside an OBS Browser Source. That gives a viewer JavaScript execution inside OBS’s embedded Chromium browser. The latest release of OBS at the time shipped a Chromium build that ran without its normal sandbox, and its V8 version was still vulnerable to `CVE-2024-7971`, a bug already…
6 min · 1,461 words
I asked Meta’s Muse for its filesystem and it sent me 6.8 GB
A security researcher asks Meta’s privileged Muse AI assistant to export its runtime filesystem—and receives a 6.8 GB dump that reveals how Muse is wired, what it can reach, and why that matters.
7 min · 1,501 words
Let the model talk. Don't let it touch the money.
Destiny Ezenwata on the hard boundary in CreditWithBleon: the LLM may converse freely, but money-moving steps stay in deterministic code—and why that line has held in production.
8 min · 1,743 words
Self-hosting LLM models for software development
Kévin Maschtaler on running medium-sized open LLMs on AWS Spot EC2 for day-to-day software work—what stacks, costs, and performance looked like versus a personal Claude subscription.
7 min · 1,719 words
JetBrains Air: Building a System of Products for Agentic Software Development
Kirill Skrygan introduces JetBrains Air: a system of products for agentic software development that treats organizational correctness—not just code generation—as the hard problem after six months of public experiments.
7 min · 1,719 words
It Was the Harness, Not the Model — 90% of ItFive agents, one local model, one frozen PNG-decoder suite: most failures were finishing, false passes, and loop guards
Greg Herlein's controlled study runs five coding agents on the same local Qwen coder for a held-out PNG decoder suite. ~90% of failures were harness problems (turn caps, early 'done', false-pass self-tests); a bigger quantization fixed none of them.
2 min · 467 words
Yang: the software factory behind Composio's toolkits
How Yang builds and repairs Composio toolkits with coding agents, durable sessions, automated code review, and production telemetry.
7 min · 1,714 words
Protobuf, JSON Schema, and OpenAPI
Buf explains how Protobuf schemas can drive JSON Schema and OpenAPI via protoc plugins—extending one source of truth into documentation, validation, and HTTP APIs without maintaining parallel definitions.
6 min · 1,373 words
Tailscale performance updates cut memory use, raise throughput, and speed startup via multi-queue, writev, and netmap caching—how the team measured and shipped the gains.
7 min · 1,640 words
How we made claude.ai 3x faster in two weeks
Anthropic’s performance sprint cut claude.ai and desktop p75 time-to-typeable from 3.1s to 0.55s: Claude Tag measured journeys, built benchmarks, and shipped thousands of guarded changes in Slack-driven loops.
18 min · 4,175 words
Microsoft killed FoxPro in 2007. Anyway, here's FoxPro revived
FoxDev Studio rebuilds Visual FoxPro as a modern 64-bit runtime and IDE in Rust/WebAssembly: open old projects in place, keep .dbf data, bridge 32-bit libraries, and add FoxScript lambdas and HTTP routes.
3 min · 645 words
PlayTaurus argues guitarists should buy hardware for what it is and run community firmware on it—ESP32-class chips are cheap enough, NAM models are open, and closed pedal firmware is the last proprietary lock.
2 min · 477 words