Practical AI Guides & a Free Claude Code Course

TimesFM 3.0 is the latest version of Google Research's pretrained time-series foundation model for forecasting. You give it a history of values — daily revenue, site traffic, stock levels — and it returns the continuation of the series together with quantile ranges, so you see not one number but a spread of likely outcomes. It was pretrained on large collections of time series, including Wikipedia pageviews and Google Trends queries, plus synthetic data. This repository holds the official PyTorch weights; the model has about 330 million parameters. The weights are released under the TimesFM Non-Commercial License v1.0.

MadsLorentzen/ai-job-search

AI Job Search is an open-source job application framework built on Claude Code. You fill in your profile once, and then a set of commands takes over: /scrape searches job portals and presents matches with fit ratings, /apply evaluates a posting, drafts a tailored CV and cover letter and has a reviewer agent critique them before the final version, and /interview helps you prepare. The author, geophysicist Mads Lorentzen, built it for his own search after his position was cut in late 2025; he reports sixty-nine tailored applications, twenty first interviews and one signed contract, and started as an AI engineer in June 2026. The core workflow is language- and country-agnostic, and other agents such as Codex or Gemini CLI can use it through AGENTS.md. It has about 42,000 GitHub stars and is MIT-licensed. It is an independent project and is not affiliated with Anthropic, the maker of Claude Code.

google/artemis

ARTEMIS is an open-source tool from Google that turns natural-language instructions into Android automation on a real phone or an emulator. It runs end-to-end workflows across apps — both testing scenarios and everyday tasks like setting up a route in Google Maps and then playing a song on YouTube. A built-in MCP server connects the phone to AI coding assistants such as Claude Code, Codex, Antigravity and Windsurf, so the agent can drive the device and collect Logcat output and screenshots. A one-click start script installs the needed tools (ADB, scrcpy, FFmpeg) and opens a local web page with a device connection wizard, live screen mirroring and execution replays. According to the project, it completes more than 99% of tasks on the AndroidWorld benchmark. It is released under the Apache-2.0 license and passed 3,000 GitHub stars in its first month.

Mistral Medium 3.5 is Mistral AI's first 'merged' flagship: a 128-billion-parameter dense model that combines instruction-following, multi-step reasoning, and code generation in a single set of open weights. Previous Mistral models were specialized — Devstral for code, Mistral Large for dialogue. Medium 3.5 does all three in one request, plus it understands images via a custom-trained vision encoder. Context window is 256,000 tokens. On SWE-Bench Verified (real GitHub bug-fixing tasks) it scores 77.6%; on the agentic τ³-Telecom benchmark — 91.4%. Reasoning effort is configurable: it can switch off for fast replies and engage fully for complex problems. Weights are open under a modified MIT license; self-hosting requires approximately four GPUs.

petergyang/no-ai-slop

No AI Slop is a writing skill that detects and removes common AI-generated text patterns while preserving the author's authentic voice. It identifies 20+ recurring patterns such as binary contrasts ('It's not X, it's Y'), throat-clearing openers, weasel attribution phrases ('experts agree'), fake-profound statements, and buzzword-heavy empty analysis. Three modes: edit mode (cleans the text and lists exactly what changed), detection mode (finds patterns without modifying the text), and satire mode (intentionally generates clichéd AI writing, for humor). Available as a skill in Claude Code and Codex via a single npx command, as a plugin in ChatGPT, or by pasting the install prompt directly into your AI chat. Install takes one command.

bytedance/deer-flow

DeerFlow (Deep Exploration and Efficient Research Flow) is an open-source super agent harness from ByteDance, the company behind TikTok. Version 2.0, launched in February 2026, rebuilt the system from scratch as a multi-agent coordination platform. Sub-agents work in parallel — one searches the web, one writes and executes code in an isolated sandbox, one reviews results — all sharing long-term memory so context survives between sessions. Built-in skills include Claude Code, Codex CLI, MCP servers, web search via Tavily, Jina, or InfoQuest, and file creation. Compatible with OpenAI, Anthropic, Gemini, DeepSeek, Qwen, and any OpenAI-compatible API including OpenRouter. A web interface opens in your browser; setup uses an interactive wizard that takes about 2 minutes. 82 000+ GitHub stars; reached #1 on GitHub Trending after version 2 launched.