AI page
Repos people are talking about
Ranked by growth inside the window: stars and forks gained, percentage growth, mentions, and how recently the code was pushed. Size alone counts for nothing. Each repo read and explained by AI.
- 81
Laya-CoreML runs the Laya typed-decision model entirely on-device on Apple Silicon via Core ML and the Neural Engine, returning calibrated probabilities for multiple-choice, score, and boolean questions with no token generation. It includes a local Snake game demo, a Python API, and reproducible latency and energy benchmarks (~5 ms per short decision on M3 Max).
Why now: It's trending as a newly popular repo showcasing validated on-device inference with notable energy-efficiency wins over MLX, though the authors note the claimed 10x speedup was not achieved.
- 82
awesome-jev is a curated list of public projects, integrations, and discussions that use Jev — TypeSafe AI's 'System One' model that turns unstructured state plus a typed question into a typed decision (choice, score, or boolean) with a confidence value. The list organizes real-world usage by category and application domain, with strict inclusion rules and explicit warnings that listing is not endorsement.
Why now: It appeared among the most-starred new repos, likely riding interest in the Jev launch and its typed-decision approach; it also explicitly addresses controversy around bulk, low-quality submissions.
- 83
Bespoke Nimble is an open recipe for building a fast, typed classification model. Given text and a schema of questions (multiple-choice or true/false), Bespoke-Nimble-9B returns each answer with per-choice probabilities in a single step, no chain-of-thought reasoning. The repo covers contrastive data curation, LoRA fine-tuning of Qwen3.5-9B, and serving, runnable on Apple Silicon or NVIDIA GPUs.
Why now: It's trending as a new, highly starred open replication of TypeSafe's Jev System One model, with a recently released 9B checkpoint, public training data, and a shared benchmark suite.
- 84
NanoJev is a small open-source Python replica of Jev, a 0.6B 'parallel decision' model. Instead of decoding tokens, it takes game states and questions and directly outputs probability distributions over 2–255 candidate choices. The repo includes an end-to-end training pipeline, datasets, and a single checkpoint that plays ViZDoom, Maze, and Snake.
Why now: A recent update (Sept 20, 2026) showed one checkpoint beating the original Jev on ViZDoom benchmarks, with side-by-side browser replays comparing NanoJev, Jev, and untuned Qwen — driving attention on trending repos.
- 85
A gallery of example apps for GeaStack, a framework for building native apps in TypeScript and CSS. Each small app (TSX) declares a manifest in package.json and can target the web simulator, ESP32 embedded boards, GeaOS, and Apple targets. The repo also includes helper tools and docs on adding and testing examples.
Why now: It was mentioned on Hacker News in discussion of native apps written in TypeScript and CSS, highlighting the unusual claim that one TSX codebase runs on web, embedded boards, and desktop OS targets.
- 86
jevchat is a fun Python experiment that abuses a Jev API to build a crude chatbot: at each step it asks the model to score candidate next symbols (from configurable alphabets like characters, words, or BPE tokens), then samples from the resulting distribution until a STOP option is drawn. It supports multiple sampling strategies, live stats, and is explicitly slow, impractical, and for laughs.
Why now: It was featured in a Show HN post ('I turned Jev into a (lousy) chatbot') as a Claude-written hack showing an absurd way to coerce a non-chat API into text generation.
- 87
Personal budget-analysis tooling for the Buckets budgeting app. It includes an interactive pie-chart explorer that reads the Buckets SQLite database (runnable as a local web app), plus scripts to merge Venmo CSV exports into a Buckets-compatible file and convert UsageDirect app-usage time into per-app cost rows. Everything runs locally so financial data stays on your machine.
Why now: It was shared on Hacker News via a 'Show HN' post about calculating how long you need to work before you can coast or retire, drawing attention to the author's budgeting tools.
- 88
Blender Copilot is a Blender add-on that embeds a Copilot-style AI chat panel directly inside Blender. An agent loop runs in Blender's own Python process, executing LLM-generated bpy code against the live scene, with each agent turn collapsed into a single Ctrl+Z undo step. It works with OpenAI-compatible providers like DeepSeek, and the README demos building an animated spaceship from five prompts.
Why now: It was just shown on Hacker News as a 'Show HN' demo, featuring a GIF of a spaceship built from five prompts inside Blender.
- 89
A Rust command-line tool that queries your local WeChat data: sessions, chat history, contacts, group members, favorites, and stats. It decrypts the local database in real time (no full pre-decryption), caches it via a background daemon for millisecond responses, can export data, and stays fully local. It also exposes agent-friendly output so AI coding assistants like Claude Code or Cursor can call it as a skill.
Why now: It appeared as a fresh, self-described restart of wx-cli and quickly gained ~740 stars on trending lists; also notable that the actual source repo (botiverse/wx-cli) is private, which the README flags as a caveat.
- 90
LocalJev is a local server, written in TypeScript for Bun, that exposes the Jev POST /v1/systemone typed decision API backed by DiffusionGemma through any OpenAI-compatible chat endpoint like oMLX. It converts typed questions into classification prompts, gets JSON probabilities from the model, retries malformed output, and returns Jev-compatible choices, scores, and confidence.
Why now: It just appeared among newly starred repos; the README contrasts it with OpenJev, which needs unmerged vLLM extensions, positioning LocalJev as a portable, wire-compatible alternative — though with self-reported rather than logit-read probabilities.
- 91
page-mascot is a React component that adds an interactive character to a webpage: it watches your cursor by picking a cell from a 3×3 sprite sheet of head directions, and shows a reaction sprite for half a second when clicked. It ships via npm, respects reduced-motion, and includes an agent skill that can draw custom mascots using AI image tools.
Why now: It is trending on GitHub as one of the most-starred new repos, likely due to its playful demo and the novel agent skill that draws and aligns custom mascot sprite sheets automatically.
- 92
Jevlike is a Python project that trains small models that pick one option from a variable-length list of text options in a single forward pass, instead of generating answers token by token. It independently reproduces the input/output shape of TypeSafe's commercial Jev model, providing training, evaluation, and prediction tools, plus demo checkpoints that score game-controller options (Doom buttons, chess keys) from image patches.
Why now: It's trending on GitHub's most-starred new repos, riding interest in TypeSafe's recently announced 'System One' Jev models whose design was never published, so this open starter implementation fills that gap.
- 93
Open Glean is an open-source Next.js AI workspace built on Hydra DB, acting as a 'second brain'. You ask questions across your memories, files, and connected apps; it retrieves context, streams answers with inline citations, and offers multi-step Deep Research, collections, mindmaps, integrations, and any OpenAI-compatible model. Keys stay server-side in encrypted cookies, and it can be self-hosted via Docker.
Why now: It appeared among the most-starred new repos, likely drawing attention as an open-source alternative to commercial AI knowledge workspaces like Glean.
- 94
Lifeboat lets you run open-weight language models on your own hardware—laptops, GPU servers, edge devices, or clusters—behind an OpenAI- and Anthropic-compatible API, with a built-in load balancer and web console. It supports pip, Docker, Kubernetes, and desktop installs, targets edge hardware like Raspberry Pi and Jetson, and claims significant throughput wins over llama.cpp and ollama. This repo is downloads and install instructions only; the source is not public.
Why now: It was featured in a Show HN post claiming 2–6x more concurrent agent sessions per GPU without quantization, plus published benchmarks showing ~2x throughput vs llama.cpp and ollama on the same hardware.
- 95
WaveDigger is a Next.js web app that looks up the physical location of Wi-Fi access points by BSSID (MAC address) and LTE/5G cell towers by network parameters, using Apple's undocumented WLOC location API. It visualizes results on an interactive deck.gl map, validates input formats, and keeps search history.
Why now: It was featured on Hacker News as a tool that digs into wireless signals to reveal physical locations, building on recent reverse-engineering of Apple's location services.
- 96
InstinctFlash is an open-source C++/Python serving runtime that accelerates robotics model inference. It supports eight model families (VLA, pi05, GR00T, Cosmos, DreamZero, etc.), applies FP8 and sampling optimizations, and reports large speedups over PyTorch on Jetson Thor and consumer RTX 4090/5090 GPUs, with a one-command `serve` entry point for fine-tuned checkpoints.
Why now: Launched via a Show HN post alongside a full-source release and fresh Jetson Thor benchmarks claiming up to 33.78× speedups.
- 97
A free, minimal macOS app (Swift, Apple silicon only) for reading Markdown files instead of editing them. It renders headings, tables, code blocks, task lists, footnotes, and offline Mermaid diagrams, with find, navigation between linked files, auto-reload, and no accounts, tracking, or configuration—well suited to reading the reports and notes AI tools generate.
Why now: It was just launched as a 'Show HN' demo of a fast, offline Markdown viewer for macOS, with the angle of reading AI/coding-agent output files.
- 98
jevmem is a TypeScript tool that automatically saves decisions, constraints, bugs, and todos from your Claude Code, Cursor, or Codex chats into a JEVMEM.md file, then injects the relevant lines back into context in later sessions. Old decisions are marked superseded rather than deleted, and lines from other sources are checked before recall.
Why now: A v0.5 release just landed, adding Claude Code plugin install, a memory-poisoning check on recalled lines, async background saving, and imports for existing CLAUDE.md/AGENTS.md files; it was also mentioned on Hacker News.
- 99
Neomacs is a work-in-progress hard fork of GNU Emacs that replaces the ~300,000-line C core with Rust and renders the UI on the GPU via wgpu. It aims for 100% Emacs Lisp compatibility, multi-threaded Elisp, 10x performance, and a zero-pause concurrent GC, using GNU Emacs itself as a test oracle to verify identical behavior.
Why now: It was featured on Hacker News as a 'GPU rendered Rust hard fork of Emacs,' sparking discussion about reimplanting Emacs in Rust while preserving compatibility.
- 100
This is the official Git mirror of WordPress core's development repository, synced from WordPress's Subversion. It lets developers fork, clone, run a Docker-based local WordPress dev environment with npm commands, run tests, and contribute patches. Pull requests must reference an existing ticket on WordPress core Trac.
Why now: It was mentioned on Hacker News in connection with a reported unauthenticated path traversal vulnerability in WordPress that can lead to conditional remote code execution.
- 101
graf is a Rust rewrite of Graphify, a tool that builds a code-graph of your codebase so coding agents can query who calls what, what depends on what, and what a change might break. It keeps the graph in SQLite, supports incremental updates, and ships as a single native binary with no Python, API key, or background service.
Why now: It was just launched on Hacker News as 'Show HN: graf (1000x faster graphify in Rust)', claiming dramatic speedups over Graphify on the VS Code repo.
- 102
Jev Review is a TypeScript code-review tool built on TypeSafe Jev. It reviews a Git diff or scans a whole codebase through a staged workflow of typed AI judgments — screening correctness, security, reliability, compatibility, and test coverage — then saves reports viewable in a local, loopback-only dashboard. Findings are meant as review prompts, not definitive defect proofs.
Why now: It's trending on GitHub's new/most-starred repos as an example of composing bounded, typed AI judgments into a practical workflow with orchestration kept in code.
- 103
Dream-RSI is a research project (paper from Google, DeepMind, UMD, UVA) on recursive self-improvement of AI agents. Its key idea is to use accumulated discovery history as a replay simulator so the agent can improve its own exploration strategy, evaluated across 8 tasks in algorithm engineering, math optimization, and GPU kernel engineering, with results like 1.22x faster downstream runtime.
Why now: It's a freshly released preprint (Sep 2026) from Google/DeepMind researchers that quickly gained ~1200 stars, and it claims progress on recursive self-improvement — a hot and controversial topic in AI research.
- 104
buried-injections is a reproducible Python benchmark that runs 10 open-source prompt-injection detectors against 629 realistic AgentDojo attacks embedded in tool output, the way agent firewalls actually see them. It shows none catches most attacks without blocking safe traffic, and that tuning thresholds dramatically reshuffles rankings — Prompt Guard 2 goes from worst to best at a 2% false-alarm budget.
Why now: It was discussed on Hacker News because its results challenge widely used prompt-injection defenses like Meta's Prompt Guard 2, showing they catch as little as 1% of realistic attacks when the default thresholds are used.
- 105
BootLife is Conway's Game of Life implemented entirely in a 512-byte x86 boot sector written in x86 assembly. Instead of RAM buffers, it uses VGA memory directly as the simulation grid, so the game runs with no operating system in BIOS real mode. You assemble it with NASM and boot the resulting image in QEMU.
Why now: It was shared on Hacker News as a 'Show HN' demo, attracting attention for squeezing a full cellular automaton into a single 512-byte boot sector.