Figma demo shows Specs and Figma agent turning context into actionable development plans
In a Workflow lab demo, Figma walks through a complete workflow: engineers use Specs to gather PRDs, Hex data, FigJam…
This category offers hands-on AI practices you can use right away: prompt writing, workflow building, tool combinations, and lessons learned. Aimed at beginners and advanced users alike, each piece gives reproducible steps and ways to verify results.
In a Workflow lab demo, Figma walks through a complete workflow: engineers use Specs to gather PRDs, Hex data, FigJam…
Kazike shared 13 plugin use cases on the 100th day since the release of DeepSeek Harness.
Runway described its practice of gradually handing each stage of software delivery over to agents. In October, agents…
OpenAI details its Performance Factory system, which uses Codex agents in the cloud to automatically investigate and…
The OpenHands team details how it hands most CVE remediation work to an agent pipeline: agents select fixes, open PRs,…
OpenAI published a long-form article detailing how the Codex sandbox runs commands within users' existing environments…
Anthropic engineer Vivek Nair reviewed the chaos loop practice ahead of Claude Docs' September 16 launch: Claude Code…
Google DeepMind's Philipp Schmid explains how to choose between Gemini 4 Hydrogen and Carbon for workloads: Carbon…
The Anthropic Claude Code team shared its verification methodology: about 90% of the team's code is written and…
The author asked an agent to compile the 10 prompts they used most often over the past two months to direct agents,…
作者拆解了近期在群里疯传的 1.4M 纯 HTML 3D 游戏,指出其中约 700KB 是打包进文件的 Three.js,其余全部由代码程序化生成,没有图片、模型和录音。
Lydia Hallie shares a Claude Code tip: use a Stop hook to make Claude retry automatically after a failed check, with…
Google DeepMind released Gemini 4 Hydrogen on November 12, 2026. On the official pricing page, the standard paid rate…
Anthropic published a blog post breaking down the architecture behind always-on agents. Such agents are stateful and…
OpenRouter ran six models, three times each, on a billing task built to fail. Without a guard, 4 of 18 runs retried…
OpenRouter has launched Beta localization in 7 languages including Simplified Chinese and Japanese, and detailed the…
Anthropic released a guide to using Claude Tag (beta) for startups, based on interviews with 8 high-growth startups…
OpenRouter released a tutorial explaining how AI agents fall into infinite loops when tools repeatedly fail, and…
Claude released a video tutorial explaining when to use Fable 5.2 and Opus 5.5 in Claude Code, when to downgrade to…
On 11 November 2026, OpenRouter published official measurements of the same GitHub task using the GitHub MCP server,…
Google DeepMind opened Gemini 4 Carbon to developers and paid users on November 10, 2026. On Google's official pricing…
OpenRouter published an explainer on quantization, covering the principles of LLM quantization and reporting tests…
OpenRouter tested the token costs of MCP, CLI and skills on the same GitHub task: on Claude Sonnet 5.5, CLI used 1.
Lovable released a security feature explainer describing how the platform helps non-technical developers build secure…
The author shares a prompt from Anthropic's Alex Albert: run the same instruction on every project, having Claude read…
OpenRouter released a tutorial explaining three ways to obtain reliable structured output from any LLM: sending a…
OpenRouter tested 15 vision models, Mistral OCR, and Tesseract on OCR performance using 8 self-made scanned pages with…
Figma officially introduced six ways to use the Figma agent on its canvas, covering stages from early design…
Anthropic's developer blog published guidance on how to choose between Haiku 5.5, Sonnet 5.5, Opus 5.5 and Fable 5.2,…
The author tested GPT-Image-2.6, released two days ago, and compiled 10 GIF prompt ideas (clay stop-motion animation,…
风险不在点过批准,而在永久规则里是否留下明文密钥。用过 Claude Code 的「Yes, and don't ask again」,可以检查本地批准规则;这不代表你已经泄密,更不代表每次批准都会永久保存。 本文依据官方文档整理,不是本…
agent-desktop is an open-source Vercel Labs desktop automation tool, not a model or an autonomous employee. An…
On NVIDIA build.nvidia.com (NIM), DeepSeek-V4.1-Flash and GLM-5.3-Flash currently expose free OpenAI-compatible API…
Bottom line: OpenAI-compatible clients (Cursor, Codex) usually set Base URL to https://host/v1. Claude Code’s…
Bottom line: there is no task-free overall winner. Shortlist by real work, then same-task test on the same materials…
[Evergreen] Bottom line: METR’s “about twice as fast” means Anthropic’s rate of capability progress on well-specified…
OpenClaw has refactored how it obtains model lists, spanning about five weeks and 80 PRs, so new models no longer wait…
Figma announced that code layers are entering open beta for 100% of users, letting teams create and edit code-based…
The author open-sourced Agent-Autoloop, an agent automation system he refined over a year, at…
Digital life creator Kazike published a DeepSeek Harness fish-raising tutorial explaining how to connect a big fat…
Sponsored
Orange Relay (pro.nat6.net) is a premium AI API relay — an excellent transfer station with competitive pricing, reliable quality and broad model coverage including Claude, GPT and Gemini.
Visit Orange Relay