設計書・コード・テストを全部AIに書かせて半年間開発してみたよ
An update on recent Claude Code quality reports \ Anthropic
Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.
Agent Harness Engineering
A coding agent is the model plus everything you build around it: prompts, tools, context policies, hooks, sandboxes, feedback loops. Harness engineering is t...
A pelican for GPT-5.5 via the semi-official Codex backdoor API
GPT-5.5 is out. It’s available in OpenAI Codex and is rolling out to paid ChatGPT subscribers. I’ve had some preview access and found it to be a fast, effective and …
Refactoring Databases in Rails
Notes from Refactoring Databases by Ambler and Sadalage, adapted for Rails. Covers drop column, rename, move column, replacing JSON with tables, and safe multi-step migration patterns.
Changes in the system prompt between Claude Opus 4.6 and 4.7
Anthropic are the only major AI lab to publish the system prompts for their user-facing chat systems. Their system prompt archive now dates all the way back to Claude 3 …
matz/spinel
Contribute to matz/spinel development by creating an account on GitHub.
Vim tip: substitute pattern delimiter doesn't need to be a slash | pawelgrzybek.com
Did you know that slash is not the only valid pattern delimiter? How many times in the past have you escaped all the slashes in the URL or file path?
GitHub Copilot個人向けプランの変更について
既存のお客様の体験を守るため、GitHub Copilot の個人向けプランに新規サインアップの一時停止・使用量制限の厳格化・モデル提供の調整を行います。変更の理由と影響をご説明します。
ENTERPILOT/GoModel: AI gateway written in Go. Lightweight unified OpenAI-compatible API for OpenAI, Anthropic, Gemini, Groq, xAI & Ollama. LiteLLM alternative with observability, guardrails, streaming, costs and usage tracking.
High-performance AI gateway written in Go - unified OpenAI-compatible API for OpenAI, Anthropic, Gemini, Groq, xAI & Ollama. LiteLLM alternative with observability, guardrails & streaming. ...
NikolayS/pgque: PgQue – Zero-bloat Postgres queue. One SQL file to install, pg_cron to tick.
PgQue – Zero-bloat Postgres queue. One SQL file to install, pg_cron to tick. - NikolayS/pgque
drasimwagan/mdv: MDV — a Markdown superset for documents, dashboards, and slides with embedded data and visualizations. HTML + PDF export, live preview, VS Code extension.
MDV — a Markdown superset for documents, dashboards, and slides with embedded data and visualizations. HTML + PDF export, live preview, VS Code extension. - drasimwagan/mdv
heygen-com/hyperframes: Write HTML. Render video. Built for agents.
Write HTML. Render video. Built for agents. Contribute to heygen-com/hyperframes development by creating an account on GitHub.
Outpost - Managed & Self-hosted infrastructure for sending webhooks by Hookdeck
Start sending webhooks in minutes at 1/10th the cost. Outpost is open-source infrastructure for delivering events to user-preferred destinations: Webhooks, SQS, RabbitMQ, Pub/Sub, EventBridge, Kafka, and more.
The Pulse: ‘Tokenmaxxing’ as a weird new trend
At Meta, Microsoft, Salesforce and other large companies, devs are purposefully burning tokens (and money!) to inflate their AI usage and hit AI usage metrics which they treat as targets.
The Rails Way in 2026
roborev-dev/roborev: Continuous background code review database for agents, work faster and smarter with accountability for every line of generated code.
Continuous background code review database for agents, work faster and smarter with accountability for every line of generated code. - roborev-dev/roborev
roborev
Continuous code review for coding agents
robzolkos/LazyPi: Pi coding agent setup for the lazy
Pi coding agent setup for the lazy
LazyPi — The fastest way to fall in love with Pi.
One command installs 60+ community skills, 76 themes, MCP servers, memory, and more.
プロダクト開発を加速させる ー ビルドトラップの向こう側へ｜migi
migi（@migii000）です。 カミナシでプロダクトマネージャー（PM/PdM）をしています。あいかわらず趣味で手書き（キーボード打ち）のブログを書いています。 さて、今回は開発プロセスのお話です。 「開発プロセスを加速させたい」と社内向けにブログを書いたら、思ったよりポジティブな反応をもらったので note で公開することにしました。 宣言をして、試行錯誤し始めた段階なので、まだ明確な成果が出ているわけではありません。しかし、なんとなくこうすると求めている成果が実現できそうだなという予感がするので、今の方針を思考のスナップショットとしてまとめておきます。 チャレンジな
Inside Notion
Company Dispatch: The culture, people, and quirks behind Notion, the first pre-GPT company to become “AI-native”
Highlights from Git 2.54
The open source Git project just released Git 2.54. Here is GitHub’s look at some of the most interesting features and changes introduced since last time.
Use Rails Combined Credentials
Rails edge (the upcoming 8.2) adds Rails.app.creds to unify ENV variables and encrypted credentials behind a single API.
snarktank/ralph: Ralph is an autonomous AI agent loop that runs repeatedly until all PRD items are complete.
Ralph is an autonomous AI agent loop that runs repeatedly until all PRD items are complete. - snarktank/ralph
earendil-works/pi-tutorial: An experimental tutorial mode for Pi
An experimental tutorial mode for Pi. Contribute to earendil-works/pi-tutorial development by creating an account on GitHub.
Haider. on X: "gpt-5.4 mini is such an underrated coding model Swe-bench pro: 5.4 high: 55.6% accuracy, $0.42 per 1m tokens 5.4 md: 55.3% accuracy, $0.29 per 1m tokens 5.4 mini-high: 52% accuracy, $0.15 per 1m tokens and significantly improves over gpt-5-mini across reasoning, multimodal https://t.co/a6gnUdjuhU" / Twitter
Swe-bench pro:
5.4 high: 55.6% accuracy, $0.42 per 1m tokens
5.4 md: 55.3% accuracy, $0.29 per 1m tokens
5.4 mini-high: 52% accuracy, $0.15 per 1m tokens
and significantly improves over gpt-5-mini across reasoning, multimodal
Tibo on X: "Hi! Getting a lot of questions on Pro plan rate limits, so I wanted to clarify a few things: - Your Pro $100 plan includes at least 10x Plus usage, till May 31 with the 2X usage boost. - Your Pro $200 plan includes at least 20x Plus usage, till May 31 with the 2X usage boost." / Twitter
- Your Pro $100 plan includes at least 10x Plus usage, till May 31 with the 2X usage boost.
- Your Pro $200 plan includes at least 20x Plus usage, till May 31 with the 2X usage boost.
Flow generation through natural language: An agentic modeling approach (2026) - Shopify
We fine-tuned Qwen3-32B into a tool-calling agent that generates Flow automations: 2.2x faster, 68% cheaper, with a weekly retraining flywheel.
microsoft/ghqr: GitHub Quick Review. Evaluate your enterprise and organizations with GitHub best practices
GitHub Quick Review. Evaluate your enterprise and organizations with GitHub best practices - microsoft/ghqr