Topic: Tools
192 stories
- Rune IDE goes open source under GPLv3, plans a revenue share for contributors
- RTK doesn't cut AI coding costs, Terminal-Bench 2.1 finds
- Perplexity trusts GPT-6 Astra with full end-to-end systems
- Pandas loses to Polars, DuckDB by up to 19x in 1-billion-row test
- OpenRouter's automatic fallbacks can change model behavior
- litelm reimplements LiteLLM's core in about 2,900 lines
- Graphify C# brings compiler-accurate Find Usages to coding agents
- Rust becomes a Tier-1 language at Microsoft, gains new MSVC compiler backend
- Gradio Workflow rebuilds AUTOMATIC1111 as a 73-node canvas
- Interactive tool renders topological surfaces as hand-inked drawings
- i-have-adhd skill stops coding agents from burying the answer
- gpu-lexer highlights syntax in any language via a WebGPU model in a 27.5KB library
- AI slop is changing how engineers review code
- Terrastruct open-sources TALA, its D2 diagram layout engine
- Mador binds DOM elements to reactive state in an 855-byte runtime
- How CronosPro's KOD cipher was cracked with the Hungarian algorithm
- Engrim gives AI coding agents a shared local memory across tools
- Asahi Linux adds official support for Apple M3 Macs
- A robotics coder gives up on letting his LLM debug visually
- VictoriaMetrics breaks down how Go's map uses Swiss Tables
- Simon Willison drives Blender with ChatGPT Codex on macOS
- Rust benchmark: cached lowercase keys beat iterator sort, unicase close
- Nitter instance tracker lists more working mirrors than before recent takedowns
- How Rust's dyn Trait builds vtables in memory
- Benedict Evans argues AI won't sweep away enterprise apps
- Balrogg losslessly recompresses Ogg Vorbis and Opus audio files
- Vite's React plugin adds native Rust React Compiler, 2.4x faster builds
- Spotify's Portal cuts Claude Code token use by 90%
- IBM markets Bob, an AI coding agent for Java and mainframe work
- Git submodules turn out to be a package manager, and a bad one
- Decompiler Explorer runs one binary through 13 decompilers at once
- Pangram's AI scores are fueling public shaming campaigns
- Nvidia launches PAIR, a free tool that links idle PCs into a home AI cluster
- Julia programming language grows from MIT project to 1M+ users
- Grep beats LSP for coding agents? HN weighs in
- Google Antigravity's terms can suspend your account for third-party use
- WebLLM runs LLMs in the browser using WebGPU acceleration
- ImHex creator reverse engineers FEZ's save file format
- Gilbert + Tobin scales ChatGPT Enterprise and Codex with CEO-led governance
- Slotstream runs 104GB Qwen3.8-Flash-Next model on 48GB Macs
- Simon Willison uses GPT-5.6-Sol and Claude Code to build a GeoJSON map viewer
- Hugging Face releases 207 WebGPU kernels, 2.57x faster than ONNX Runtime Web
- Bupa uses AI to migrate legacy app, cuts delivery time 60%
- Ambient CSS v3 recreates Blender lighting in pure CSS
- Graham Dumpleton launches Wrapture, a Python testing and tracing tool built entirely by AI
- Essay: four years of AI coding tools, still no new Airbnbs
- DoltLite reaches Beta with Git-style version control for SQLite
- BirdNET-Go turns security cameras into a bird identifier
- Zig ships pointer stability locks for ArrayList, reworks packaging
- Wirewiki's autocomplete hits a 121ms budget across 240M domains
- sm750hdmifb driver brings real ultrawide output to old SM750 GPUs
- Relm4 brings Elm-style declarative UI building to Rust
- vphone-cli boots a virtual iPhone via Apple's Virtualization.framework
- Tether adds iMessage, SMS and notifications to Linux
- Open OSCAR Server revives classic AIM and ICQ chat clients
- How to run an AI chatbot on your own computer
- Defragger gives Linux a real graphical disk defragmenter
- CPython officially adds tier 3 RISC-V support
- StemDeck splits songs into six stems, running fully offline
- OpenAI's Python SDK migrates to HTTPX2, changing TLS trust store
- htmx 4.0 ships, moving to fetch() from XMLHttpRequest
- Flyway Community edition gets a rollback trick via version numbers
- Sentence Transformers adds training support for multi-vector ColBERT-style embeddings
- Experiential launches an open source gateway that turns usage into a custom router or model
- Anthropic's Model Hardware Standard lets AI agents control lab hardware
- AI Engineer Notebooks teaches applied LLM skills without frameworks
- Tailscale releases Tailcat, netcat over WireGuard tunnels
- pnpm 12.0 ships as a Rust rewrite, not a migration
- CoMaps guided earthquake rescuers offline in Venezuela with no cell signal
- Amazon shuts down Mechanical Turk on September 30, 2026
- TeXbrain compiles LaTeX to PDF in the browser via WASM
- Nitter goes dark after cease-and-desist letters, XCancel reportedly targeted too
- LatticeDB combines graph, vector and text search in one file
- A warm window pattern stops tooltip delays from stacking
- A longtime user says Claude Code and Opus 5 lost their focus
- llm-anthropic 0.27 adds anthropic v1.0.0 support
- Gradio adds gr.Workflow, a drag-and-drop AI pipeline builder
- Ambient Context turns your screen into a markdown memory for LLMs
- Adi2 brings CSS styling and XML layouts to native Ada GUIs
- Kern runs containers from a 1.52 MB binary with no daemon
- Vibe-coded apps are spawning a cleanup industry, QA founder says
- Spotfire webinar pitches agentic AI for chip yield root cause analysis
- Simon Willison ships llm 0.33 with combinable templates
- Linus Torvalds credits AI for grinding through a hellish kernel bug
- Hister ships a self-hosted full-text search index you fully control
- Figmimic turns any webpage into editable Figma layers
- Developer compares Codex and Claude Code after a week
- Stop building TUIs: AI makes native GUIs cheap now
- Rust Glancer launches as a low-memory alternative to rust-analyzer
- OzBrain launches a shared brain that AI agents read and write
- New Claude Code skill runs Claude's replies through Gemini to cut buzzwords
- Huzzah turns AI coding prompts into persistent pseudocode
- Bun 1.4 adds Bun.WebView for built-in browser automation
- Unsloth releases Dynamic v3.0 GGUFs, over 10% more accurate than other providers
- SvelteKit 3 takes on Next.js with type-safe remote functions
- Replit launches Free Mode on GPT-5.6 Luna, offering unmetered software creation
- Go 1.27 ships with generic methods and post-quantum crypto
- Turbovec brings Google's TurboQuant quantizer to Rust
- SoLo lets fully static Linux binaries load host GPU drivers
- Sentence Transformers v6.0 adds ColBERT-style embeddings
- Mojo open-sources compiler and toolchain under Apache 2
- modelmap.cc visualizes Hugging Face model architectures
- DesktopFly simulates a real fly brain from FlyWire data on macOS
- Cursor launches Origin, a GitHub rival, amid outage backlash
- Show HN: Vocal Slice cuts audio clips by selecting transcript text
- Sebastian Raschka builds an AI text detector from scratch
- Buf ships production-grade LSP server for Protobuf
- tawc brings hardware-accelerated Linux apps to Android, built largely with Claude Code
- Simon Willison builds CORS Chat, a web UI for LLM endpoints
- Sebastian Raschka builds an AI text detector and a model that avoids it
- QuantStack brings Numba JIT compiler to the browser via JupyterLite
- Artificial Analysis launches Optima to benchmark AI models on your own data
- ThoughtDAG turns LLM chat history into an editable context graph
- Simon Willison: have LLMs hallucinate tags, then match via embeddings
- RustDesk adds unattended remote access on Wayland
- Firefox is now the last major browser to support uBlock Origin
- Ember ships redshift-safe palettes for terminals and charts
- Deltix launches AI agent that tests mobile apps from plain English tasks
- Anthropic announces watermark detection API for Claude text
- llm-gemini 0.33 adds Gemini 3.7 Flash, reasoning traces
- Zed launches Delta, multiplayer coding with agents
- Woxi reimplements Wolfram Language as a Rust interpreter
- RingCentral rolls out ChatGPT Work and Codex to every employee
- OlmoEarth Studio adds custom embedding export for Earth observation
- rag-staleness-check flags stale, orphaned and duplicate RAG vectors
- Modular ships Mojo 1.0 as a stable systems language
- llama.cpp ships llama.app for one-line local model installs
- Line9 launches a Mermaid renderer with its own layout engine
- Jolt compiles Clojure to native binaries, no JVM needed
- Google argues Go is the right language for AI coding agents
- datasette-upload-dbs 0.5a0 adds a formal swap API
- World Train Map redraws 1,093 routes with OpenStreetMap data
- VectorWare brings Rust's portable SIMD to the GPU
- Sonic Pi v5 ships with overhauled interface and synth engine
- Meetily transcribes and summarizes meetings for free
- mcptoon claims 98% cut in MCP tool discovery tokens
- Juror open-sources a multi-model alternative to Greptile
- Hyperspace 1.7.1 dedupes files on macOS using space clones
- H3-metal ports MiniMax-H3 video generation to Apple Silicon
- GPT-5.6 Sol Pro compresses SQLite revision history over 250x
- Every Cube lets you scroll through all 43 quintillion Rubik's Cube states
- Klepton runs Android VR APKs on Apple Vision Pro
- GitHub retires GitHub Models, the unified LLM API for Actions
- Docker launches Sandboxes, microVM isolation for AI agents
- AI writing detectors are unreliable, but everyone uses them anyway
- Triton brings full DirectX 11 support to Windows guests in QEMU
- TheoremDB opens as a public workspace for machine mathematics
- Stylized GGX Shading blends toon and PBR specular highlights
- Microsoft Word 1.1a gets a native x64 port
- MarketNow's MCP directory lists 9,248 servers, not the interceptor its HN post claims
- A dithering trick hides a full photo inside a scannable QR code
- Wyzer merges Perceus memory management with choreographic programming
- Rippling launches AI Spend Console to rein in runaway AI spend
- HSP GRUPPE deploys ChatGPT Enterprise across 81 tax firm groups
- Zed opens early access to DeltaDB, a version control system linked to agent conversations
- Delaunay32 claims over 10x faster exact 2D triangulation
- MCP 2.0 drops session state; Willison ships three tools on it
- LLM 0.32 adds reasoning traces, server-side tools, and MCP support
- Flowise shuts down as developers move to coding agents
- Treblo's detector flags Fenix Flexin's "Rubberz" as AI-made
- Terence Tao's ChatGPT chats show why domain expertise wins
- Swiftlet runs 80B Qwen on 4.3GB of Mac RAM, 35B on iPhone
- Developer manually retypes LLM code to avoid cognitive debt
- AWS helps Superblocks embed vibe coding in private clouds
- AirLLM runs the 2.8T-parameter Kimi K3 on under 4GB of VRAM
- SwiftUI remains a perpetual beta seven years after launch, critique argues
- Stack Overflow: AI coding tool use is up, trust is down
- ssh.place turns an SSH session into a shared pixel canvas
- Simon Willison ships condense-json 1.0 for deduplicating JSON
- Shitty rewrites Zutty into a GPU-accelerated, memory-unsafe terminal
- pytest-leak-finder bisects test suites to find state-leaking tests
- Mu bundles 67 real internet services behind one MCP endpoint
- Kakehashi runs macOS ARM64 CLI binaries on Linux aarch64 natively
- Isopolis renders San Francisco as an isometric pixel-art map
- F*: the proof-oriented language behind Firefox and Linux crypto code
- Go 1.27 adds generic methods, ML-DSA post-quantum crypto and a UUID package
- Glyphs 4 ships with visual kerning and COLRv1 color fonts
- Diátaxis splits documentation into four types, already adopted by Cloudflare and Gatsby
- CostPerPrompt tracks live pricing across 232+ AI models
- WASTE engine streams full 2.78T-parameter Kimi K3 from disk
- Solid Queue 1.6.0 adds fiber workers for Rails background jobs
- Simon Willison ships llm-mcp-client alpha for stateless MCP 2.0
- Simon Willison builds three tools around stateless MCP 2.0
- pgtestdb ties schema-based Postgres tests on setup speed
- Personal software gets cheap enough to build in a weekend
- Microsoft's Flint pitches one chart spec for five rendering backends
- Cursor strips dollar costs from usage dashboard and CSV export
- TurboFieldfare runs Gemma 4 26B in 2GB of RAM on 8GB Apple Silicon Macs
- Tokenless launches AI model router that halves inference bills
- Pangram 4 claims one error per 24,000 documents in AI detection
- Kuna, an LLM-written decompiler, closes in on IDA Pro
- claude-code-merge-queue serializes landings for parallel Claude Code agents