Two defaults changed while you weren't looking, plus a one-click DeepSeek Harness
Two tools most of us stopped reading release notes for changed this year, and neither made much noise about it. Transformers, the library behind almost every AI tutorial you have read, dropped TensorFlow and JAX in version 5. It is PyTorch only now, so a lot of older example code no longer runs as written. VS Code folded Copilot Chat into the main repo as open source and built AI into the editor itself, with a single setting that turns all of it off. When roughly three out of four developers use the same editor, that setting is worth knowing about. Defaults get to do that. Once a tool is the thing nobody has to justify choosing, it can change underneath you and you find out from a broken tutorial. The other two picks are trying to earn that position, and both are doing it the same way: by removing the setup. DSH Desktop, this week's featured tool, wraps DeepSeek's agent runtime in a plain installer for Windows and Mac, so trying it no longer starts with a terminal. It is a community project, not DeepSeek's, and it is free. Understand-Anything does the same for your own code. It maps a repo into a knowledge graph you can click through and question, then saves the graph as JSON so the next person skips the expensive part. One honest warning on the featured pick. DSH Desktop pins a release candidate of a runtime that calls itself a developer preview and promises breaking changes in capital letters. Install it to learn the runtime. Keep it off anything with a deadline.
A community-built desktop app for DeepSeek Harness, with the plugin market and mobile remote control built in. Free, MIT, and not an official DeepSeek project.
The Lens
DSH Desktop turns DeepSeek Harness into a normal app. The harness is DeepSeek's open source agent runtime, and running it means Node, pnpm, a terminal, and a port. This is one installer for Windows x64 or any Mac, Intel or Apple Silicon, and the Web UI, host service, and plugin system open in a native window. MIT, free, and not an official DeepSeek project. Nothing to host. The installer bundles Electron, Node, pnpm, and a pinned harness build, and leaves your global PATH alone. Updates download only after you confirm. Upstream runs unmodified, so official plugins install from the bundled terminal. Newer builds add a built-in community plugin market and mobile remote control. Solo and small teams on Windows or a Mac should use this instead of the npx command. Medium and large teams get nothing extra: no shared profiles, no policy controls. Anyone already running the harness in a browser gets more from zhu1090093659/dsh-web-ui. For a multi-provider chat client instead, CherryHQ/cherry-studio is further along. The catch is what it pins. Upstream calls itself a developer preview and warns in capitals that breaking changes are coming, and this ships a fixed release candidate. No Linux installer has shipped yet, the issue tracker carries launch crashes on both supported platforms, and LAN mode opens the UI to your network with no authentication.
Load almost any open AI model in a few lines of Python. The standard library for running and fine-tuning models from the Hugging Face Hub, now PyTorch only.
The Lens
Hugging Face Transformers is the standard Python library for working with pre-trained AI models: text generation, translation, summarization, image classification, speech, and more. The Hub it connects to lists over a million Transformers checkpoints. Load a model in three lines of code, run inference, done. Apache 2.0 and free, with no gated features. One change matters for anyone following older tutorials. Version 5 removed the TensorFlow and JAX backends, so the library is PyTorch only and needs Python 3.10 or later. The pipeline API still takes you from zero to a working sentiment classifier in one line, and the Trainer API handles the training loop, checkpointing, and evaluation for fine-tuning. The library is free. The Hugging Face Hub is free too, with 100 GB of private storage. Pro is $9/mo for 1 TB. Team is $20 per user per month and adds SSO and audit logs, and Enterprise is $50 per user. Solo and small teams: free library, free Hub. Medium: Team once you need SSO. Large: Enterprise, or your own GPUs. The catch: big models need serious GPU hardware. A 7B parameter model needs about 14 GB of VRAM just to load in half precision. Managed Inference Endpoints start at $0.03/hr on CPU, with a T4 GPU at $0.50/hr and an A100 at $2.50/hr. The install is heavy because it pulls in PyTorch. For production serving, vllm-project/vllm or ggml-org/llama.cpp will perform better.
Maps how a codebase fits together so you can see it instead of grepping it. Runs inside Claude Code, Cursor, Copilot, and other coding agents.
The Lens
Understand-Anything turns any codebase, knowledge base, or set of docs into an interactive knowledge graph you can search, click through, and ask questions about. It maps out components, relationships, and data flows so you can actually see how a system fits together instead of grep-ing your way through it. It runs as a multi-agent pipeline on top of your AI coding setup. Tree-sitter handles the deterministic parsing, then LLMs do the semantic analysis. Originally Claude Code only, it now works with Cursor, VS Code with Copilot, and other agent harnesses. MIT licensed, nothing to host. Graphs commit to your repo as JSON so the rest of the team gets the same map. Solo devs onboarding to a messy codebase get the biggest win. Small teams use it to document tribal knowledge that lives in someone's head. Large teams probably already have wikis and ADRs, but the graph can still surface dependencies the docs missed. The catch: graph quality tracks codebase quality. Spaghetti in produces spaghetti out, and circular dependencies render as a hairball. You also pay for the LLM tokens, and large monorepos can run up a real bill on the first pass.
The editor nobody has to justify choosing. Free, extensible for any language, and its AI chat now ships as open source.
The Lens
VS Code is the default code editor: editing, debugging, Git integration, and extensions for any language. About three quarters of developers use it, per Stack Overflow's 2025 survey. It's the editor you don't have to justify choosing. The source is MIT licensed, while the official download is a Microsoft build with its own license and telemetry. Built-in terminal, built-in Git, built-in debugging for most languages, and a marketplace with tens of thousands of extensions. The editor is free for private and commercial use, and so is the web version at vscode.dev. Most extensions are free, though publishers can charge, and the built-in AI features run on your GitHub Copilot plan or your own model key. Copilot Chat itself is now open source inside the main repo, and one setting turns all the AI features off. Solo through enterprise: free at every scale. Settings Sync keeps your setup consistent across machines. Remote development extensions let you code on a server or inside a container from your local machine. Live Share does real-time collaboration. The catch: it's an Electron app, a web browser pretending to be a desktop app, and it gets heavy with many extensions loaded. The Microsoft telemetry in the official binary bothers some people, and you can opt out of much of it but not all. VSCodium is a community build without it. For a lighter, keyboard-driven workflow, neovim/neovim or zed-industries/zed will feel faster.
Get the next issue in your inbox
Free. No spam. Unsubscribe anytime.