GitHub trends for September 23, 2026
What was taking off on GitHub on September 23, 2026: peak magnitude of the day, summary and where it was spotted.
-
8.9
jaredpalmer/kev
kev is a LoRA adapter with a small readout head on top of Qwen2.5-0.5B that answers many typed questions about a document in a single forward pass, returning calibrated probabilities instead of text.
-
8.8
unreallabsai/unreal-agent
Async-first agent harness in Go: sessions with persisted forkable history, input deduplication, an LLM-turn coordinator, and asynchronous operation execution via swappable components.
-
8.7
NandhaKishorM/laya
Laya is a non-autoregressive decision engine that answers typed questions (choice, score, noul) over any state (text, email, ticket, JSON) in a single forward pass (~35 ms on GPU) with calibrated probabilities.
-
8.6
google/ax
A declarative orchestrator for running autonomous AI agent workloads in a cluster, using a Kubernetes-like model with Task, Workspace, Gateway, and Model manifests. It sandboxes agents, pre-wires their environment, and…
-
8.3
CopilotKit/openmuse
A personal AI agent with a browser, terminal, and files that carries out tasks on request and shows its work. Built with CopilotKit and AG-UI for iOS, Android, and web.
-
8.1
jev-chat/jev-chat-jarvis
Android app that reads the on-screen chat via AccessibilityService, judges the other side's intent with the Jev model, and shows 3 ranked reply drafts in a translucent overlay. Replies are filled into the input box but…
-
7.3
browser-use/jev-ultrafast
A browser agent with a dynamic, indexed action space: given one natural-language goal it picks an operation (CLICK, TYPE_TEXT, SELECT, etc.) and an element, using a small LLM to generate text only for typing.
-
7.3
browserbase/stagehand
SDK in TypeScript, Python and Go for LLM-driven browser control: observe/act/extract find selectors, perform actions and return schema-validated data. Built for agents that operate websites like humans.
-
7.2
dream-num/univer
Open-source TypeScript SDK for embedding spreadsheet, document, and presentation editors into your own product, with a plugin architecture, Canvas rendering, a formula engine, and one Facade API for browser and…
-
7.2
Mak5er/AirCard
A macOS tool that customizes Apple Wallet cards and lockscreen passcode keypad themes on iOS 18+ without jailbreak via the airlift exploit. It lets you assign custom artwork to Apple Pay cards and apply .passthm themes.
-
7.2
volotat/mini-AGI
A byte-level continual-learning language model that assembles its own architecture, trains from scratch on a single 8 GB VRAM GPU, and keeps learning from everything it reads. An experiment showing continual learning…
-
7.1
mizorewww/laya-mlx
Native MLX runtime for Laya typed decision models: returns probabilities for choices, scores or truth values without text generation, running locally on Apple Silicon.
-
6.9
mvt-project/mvt
A collection of mobile forensics utilities that gather and analyze traces of compromise on Android and iOS devices, matching them against public indicators of compromise (IOCs). Built by Amnesty International Security…
-
6.8
anthropics/financial-services
Reference agents, skills, and MCP connectors for financial-services workflows: investment banking, equity research, private equity, and wealth management. Installs as a Claude Cowork plugin or deploys via the Claude…
-
6.8
heyjunpenn/awesome-jev
A catalog of 530 open-source projects built with Jev, TypeSafe AI's System One model for typed decisions inside software. It serves as a curated list of implementations, SDKs, and tooling for evaluating the Jev…
-
6.6
TheoLeeCJ/SemIf
An open reproduction of the Jev-style semantic decision interface: reads typed option probabilities directly from a 4B model's logits without generating text. Runs on a single RTX 3090.
-
6.4
Amal-David/awesome-jev
A curated catalog of projects, demos and SDKs for TypeSafe Jev, a model that answers typed questions about text or JSON. It helps move from a demo clip to understandable code and separates reviewed sources from the…
-
6.4
freestylefly/WeChatBridge
A native macOS tool that uses a Share Extension to take a WeChat chat export and forward it to AI agents (Codex, Claude, Doubao, etc.) or archive it into Obsidian. Runs fully locally without reading the WeChat database.
-
6.2
hydra-db/open-glean
An open-source AI workspace over Hydra DB: it answers questions across your files, notes and connected apps, writes the answer and cites its sources.
-
6.1
Open-Dev-Society/OpenStock
An open-source alternative to paid market platforms: real-time price tracking, personalized alerts, and company insights. Built with Next.js, MongoDB, and Finnhub, free for everyone.
-
6.0
jev-chat/jev-chat-windows
A reply helper for WeChat on Windows: it captures the window, reads incoming messages with offline OCR, infers the other person's intent, and offers 3 candidate replies to insert into the input box. Sending is always…
-
6.0
malevrigns/agent-jev
AgentJev-0.6B is a specialist decision model for AI agent loops: it takes unstructured state (diffs, traces, logs) plus typed questions and returns a probability distribution over options in a single forward pass with…
-
6.0
miuuyy/Astra-Ares
Experimental reference tool that dynamically selects reasoning effort for GPT-6 Astra via Jev during Codex tasks to reduce token usage. Requires building a patched Codex CLI from source.
-
5.9
mizorewww/laya-coreml
Local port of the Laya model to Apple Core ML and Neural Engine: returns typed decisions (choice, score, yes/no) without token generation, with speed and energy benchmarks.
-
5.8
robbietilton/Compositor
A free, open-source raster image editor for macOS written in Swift — a Photoshop alternative focused on compositing and post-processing.
-
5.8
typesafe-ai/skills
Agent skills for building with TypeSafe's System One API: typed decisions and probabilities for LLM agents. Installs as a Claude Code plugin or via skills.sh.
-
5.7
maanHimself/OpenDLSS-NR
A Vulkan reimplementation of NVIDIA's DLSS 5 Neural Rendering network, bit-exact against the original. Requires your own weights and an NVIDIA Ada or newer GPU.
-
5.7
trycua/cua
Open-source toolkit for computer-use agents: cloud and local desktops, an app automation driver, and benchmarks for training and evaluation. For those giving AI agents control of a computer.
-
5.7
yibie/awesome-jev
A curated list of public projects, integrations, and discussions built on Jev — TypeSafe AI's System One model for typed decisions (choice, score, boolean with confidence).
-
5.7
zai-org/ZCode
Z.ai's coding agent workbench: an Electron desktop app, browser UI and terminal Agent CLI. The repo contains the client, backend, shared UI and the agent CLI/runtime sources.
-
5.6
BuilderIO/agent-native
An open-source TypeScript framework for building agentic apps where each capability is defined once as an action, used by the agent as a tool and called by the UI from code. Aimed at developers who need agents paired…
-
5.6
jev-chat/jev-chat-jarvis-mac
A macOS app that OCR-reads WeChat messages on screen and uses a local model to classify intent and risk, suggesting reply candidates. Read-only, with no injection into WeChat.
-
5.6
newliver666/apk-reverse
An Agent Skill for Android APK reverse engineering: unpacking, ad removal, dex patching, repacking, and runtime/server analysis. Loaded by an agent (Claude Code, Codex) while it works.
-
5.5
agent-substrate/substrate
Agent Substrate is a secure-by-default agent execution runtime built on Kubernetes that multiplexes many "actors" onto a small pool of workers with sub-second suspend/resume and isolation via microVMs and gVisor. It is…
-
5.4
MelisaPeteRs2006/Silent-Crypto-Miner
A cryptocurrency mining builder and runtime with CPU/GPU support, multiple algorithms and pools, plus a stealth mode (process disguise, auto-start, pause on activity).
-
5.3
bendlang/bend
Bend 2 is a dependently-typed programming language with proofs that compiles to CPU and GPU. It lets you declare laws in LAWS.bend that the compiler mechanically verifies, blocking AI-agent mistakes.
-
5.3
dorkitude/webctl
A web search CLI for AI agents that filters and deduplicates search results and page chunks via Jev, saving context tokens.
-
5.3
latent-spaces/brag
An agent skill (Claude Code, Codex, opencode, etc.) that turns a project you just built into a short launch video with music, motion, and share copy — in one command. Rendering is powered by Hyperframes.
-
5.3
Tencent/BrowserSkill
BrowserSkill connects AI agents (Cursor, Claude Code, Codex, etc.) to your already logged-in browser via the bsk CLI and an extension, letting the agent work in a separate window without interrupting you.
-
5.2
anmolkapil/plexo
Download manager for Windows and macOS that speeds up downloads by splitting files into 8 MB chunks fetched in parallel across multiple network interfaces (Wi-Fi, Ethernet, USB tethering).
-
5.2
davila7/claude-code-templates
CLI tool for installing and configuring ready-made Claude Code setups: agents, slash commands, MCP integrations, hooks and settings, plus session monitoring. Aimed at developers using Claude Code.
-
5.2
geastack/examples
A collection of example applications and tools for GeaStack: small TSX apps with a gea manifest, used by the simulator, embedded boards, GeaOS and the IDE extension.
-
5.2
jarrodwatts/jev-trader
Trading bot for the Kuru DEX on Monad: every block (~300 ms) an AI model decides buy or sell from the MON-USDC order book and posts a post-only limit order one tick inside the touch, earning the spread. Includes a…
-
5.2
kerpopule/hermes-jev-skills
A set of eight SKILL.md files and a CLI that wires the Jev decision model into Hermes, Claude Code and Codex agents for model routing, memory filtering, skill selection and GUI actions.
-
5.2
lhlGitHub/threejs-architecture-effects
An Agent Skill for Codex, Claude Code and Cursor that generates an interactive Vite + React + Three.js site with a procedural 3D building assembled along a 0–1 timeline.
-
5.2
lordx64/phantom-kv
Removes refusal behavior from language models by grafting a small trained KV-cache bank into the context, without editing weights or projecting activations, and is fully reversible.
-
5.2
QwenLM/Qwen-Image-2.1
Qwen's open-source text-to-image generation and image editing model, including transparent RGBA output and up to 10 reference images. 7B generation component, 2K support, integrations with Diffusers, ComfyUI, vLLM,…
-
5.2
Rizzo-AI-Academy/rizzo-flow
A local-first implementation of the Jev idea: an LLM returns typed decisions (boolean, choice, score, numeric) with probabilities in a single forward pass, without generating tokens. Compatible with the TypeSafe HTTP…
-
5.2
ruanyf/weekly
A weekly tech digest by Ruan Yifeng: articles, tools and resources, published every Friday. Also includes a section with programmer job postings.
-
5.0
FluidInference/FluidUse
Local computer use on Apple silicon: reads forms in macOS apps via the Accessibility API, asks a small on-device model what belongs in each field, and types the answer into the real app.
-
5.0
hypit-ai/hypit
Hypit gives AI agents a language and system to create video: the agent clones a viral clip as an editable workflow — footage, captions, B-roll and effects anchored to words instead of seconds.
-
5.0
Mak5er/AirCard-iOS
iOS app that writes custom artwork into iOS 27 system caches over a local loopback tunnel, letting you restyle Apple Wallet cards and the lock screen passcode keypad directly on the iPhone.
-
5.0
Nasiko-Labs/nasiko
A single control-plane for AI agents: deploy, route, proxy A2A traffic, plus MCP gateway, LLM router and tracing in one Rust process.
-
5.0
receptron/laya
TypeScript wrapper for running the open-source Laya System-1 decision model via ONNX Runtime: takes a state and typed questions, returns answers with calibrated probabilities in one forward pass.
-
5.0
strands-agents/harness-sdk
Open-source Python and TypeScript SDK for building and running AI agents: agent loop, tools, MCP, memory, multi-agent patterns and tracing in one package. Runs in your process with no hosted control plane and supports…
-
5.0
tamaratran/fast-jev-compaction
Claude Code plugin that replaces context compaction summaries with Jev decisions: every tool call and result is scored in one fast request, stale ones are dropped or truncated, everything kept stays verbatim. Also…
-
5.0
wfzyx/von
An open-source non-autoregressive System One decision model for classification, intent routing, and guardrail validation without text generation, with calibrated probabilities and sub-25ms latency.
-
4.9
cloudflare/security-audit-skill
A coding-agent skill that turns the agent into a security auditor, running a multi-phase code audit with independently verified, machine-readable findings.
-
4.9
laoma2053/awesome-zhuiju-free
A curated list of free movie and TV streaming resources: online sites, apps, cloud and torrent search, subtitles, TVBox configs, IPTV and related open-source projects. Resources are hand-picked and checked daily for…
-
4.9
zhouxiaoka/autoclip
AI-powered video clipping tool that downloads videos from YouTube and Bilibili, analyzes content with Qwen LLM, and automatically extracts highlight clips and generates compilations.
-
4.8
hydra-db/hydradb
HydraDB is a distributed graph database written in Rust that stores the graph durably on S3-compatible object storage, with separate data nodes and indexers. It supports OpenCypher queries, GraphBLAS traversal, Neo4j…
-
4.8
Lumid-Off/AirCard-Windows
Native Windows app in Rust for customizing Apple Wallet cards and lockscreen passcode themes on iOS 18+ without jailbreak via the AirTraffic sync exploit.
-
4.8
rgem227/knoweldge-base
A self-hosted personal or team file knowledge base built with FastAPI and SQLite, with folders, tags, in-app previews, and a read-only MCP service for agents.
-
4.8
superdesigndev/treg
Treg is a proxy registry of tools for AI agents: one base URL and token give access to 3,000+ endpoints across 60+ providers (SEO, social, data enrichment, media generation) with per-call pricing and no provider signup.
-
4.7
browser-use/video-use
A skill that lets coding agents (Claude Code, Codex) edit raw video footage via chat: it transcribes takes with ElevenLabs, builds an edit decision list, and renders final.mp4 with cuts, color grades, subtitles and…
-
4.7
ipenywis/laya-ultrafast
A local port of the jev-ultrafast browser agent that makes action decisions with the open-weight Laya model via MLX instead of a hosted API. Apple Silicon only, and it runs browser tasks without per-step cost.
-
4.7
stablyai/orca
Orca is an ADE for running a fleet of parallel coding agents (Claude Code, Codex, OpenCode, etc.) in isolated git worktrees with a unified interface, mobile companion, and remote runtime.
-
4.7
Zefan-Cai/Open-Jev
Open-Jev serves a context, questions and candidate answers to Qwen-based models and returns typed probabilities directly, without autoregressive generation or JSON parsing. It provides LoRA adapters, a scalar decision…
-
4.6
CTNicholas/jev-workflow-builder
A TypeScript demo app for building multiplayer workflows that wire together Jev and LLMs using Liveblocks. It lets you assemble call chains, invoke them via REST API, and see other participants' state and cursors.
-
4.6
Driftmireminisce/MarkText-Pro-Desktop
A lightweight desktop companion for MarkText: a Markdown editor for notes and writing with dark and light modes and easy customization. Builds are available in the Releases tab.
-
4.6
ethanplusai/astra-flash-orchestrator
A Codex skill that splits coding work between two models: Astra plans, reviews and makes decisions, while DeepSeek V4.1 Flash handles bulk implementation, testing and debugging. Aims to cut the cost of expensive model…
-
4.6
haibarazz/awesome-codex-research
A collection of reusable Skills and Codex Plugins for research workflows: literature review, paper polishing, review-comment decomposition, ML experiment planning, and remote GPU control via MCP. Aimed at researchers…
-
4.6
jerryjliu/docjev
Python library, CLI, and local app for classifying documents into categories and splitting PDF/DOCX/PPTX packets into separate documents using natural-language rules.
-
4.6
vercel-labs/json-render
Generative UI framework: an LLM turns prompts into JSON interfaces constrained to a catalog of components you define, keeping output predictable and safe. Supports React, Vue, Svelte, Solid, React Native, Next.js, 3D,…
-
4.6
wdobry/laya-playground
A playground for Laya, the open-source decision model: a website, three games, a benchmark and an agent skill. A local server takes typed questions and returns probabilities in one forward pass.
-
4.6
yynxxxxx/Codex-X
Cross-platform GUI tool for OpenAI Codex (desktop/CLI): manage providers and APIs, prompt templates, sessions, Skills/MCP and TOML config without hand-editing files.
-
4.5
BansheeLight/MongoDB-Suite
A lightweight local-first companion tool for everyday MongoDB workflows, based on the repository topics (backend, database, nosql). The README gives few details beyond customization, small footprint, and local-first…
-
4.5
Michael-Jiahao-Zhang/game-the-llm-reviewer
An agent skill that performs a final editing pass on research papers, rewording passages to match documented LLM reviewer preferences while preserving scientific meaning.
-
4.5
Re1nhard-1/deconv-preflight
A Python CLI that validates bulk expression tables and single-cell reference metadata before deconvolution, catching missing or duplicate IDs, cell mismatches, and zero-only samples.
-
4.5
TianyuCodings/NanoJev
A nano replica of Jev built on Qwen3-0.6B that outputs probability distributions over candidates in a single forward pass with no token decoding, plus a training pipeline and game demos (maze, Snake).
-
4.4
blixvip/NullMotion
A web tool that plays a finished motion-graphics ad over its black-and-white HyperFrames drafts, keeping them in sync with the film and letting you export everything to MP4 frame by frame.
-
4.4
cloudflare/quiche
A Rust implementation of the QUIC transport protocol and HTTP/3 with a low-level API: the application handles sockets, event loop and timers itself. Used in Cloudflare's edge network, Android's DNS resolver and curl…
-
4.4
deepseek-ai/deepseek-harness
Open-source agent harness by DeepSeek AI written in TypeScript, built on an everything-is-a-plugin architecture powered by Cordis. It runs as a web UI via npx and serves as a base for composing AI agents from plugins.
-
4.4
eternity4719/HowToLiveBetter
An evidence-based life guide: 498 recommendations on longevity, first aid, personal finance, law, employment, family and emigration. Each lists costs, benefits, evidence level and primary sources.
-
4.4
TencentARC/GameHorizon
Official repo of GameHorizon, a data and evaluation suite for measuring model gameplay capabilities in AAA games across different temporal horizons, with 5,000 hours of gameplay and offline/online benchmarks.
-
4.4
vinnylarouge/jevlike
A small trainable model that picks one option from a changing list of text options in a single forward pass, an open alternative to TypeSafe's Jev. Aimed at developers who need fast classification or menu selection…
-
4.3
affaan-m/ECC
A performance optimization system for coding agents: a set of agents, skills, memory, and rules that adds a plan-test-review-remember workflow to Claude Code, Codex, Cursor, and other harnesses.
-
4.3
awlevin/typesafe-computer-use
Drives macOS toward a goal typed in plain English: OCR the screen, pick the next action via a TypeSafe classifier, click. Costs about $0.0002 per step and never sends screenshots to large models.
-
4.3
bespokelabsai/nimble
Nimble is a model and training recipe for fast typed decisions over text: given a flat schema of enum and boolean fields, it picks an answer and returns probabilities for each option. It targets routing, condition…
-
4.3
cdyforever/how-to-live-better
A single-file HTML reading page for the open book "High Cost-Performance Life Guide" — 528 tips across 32 sections, searchable, offline-capable, zero dependencies.
-
4.3
deepopen-com/deepopen
Open-source non-autoregressive System 1 decision engine for typed decisions: returns typed answers (choice, score, noul) in a single forward pass across 100+ languages without generating text. Suited for…
-
4.3
vinilana/jev-gateway
A local LLM gateway for coding agents: before choosing a tool, the agent asks the fast Jev model which tool to call instead of the expensive reasoning model. Works with Codex, Claude Code, OpenCode and Gemini CLI.
-
4.2
akitaonrails/ai-memory
Long-term memory server for AI coding agents: stores notes as a markdown wiki and hands off context between different agents and machines. Lets you resume a task in another CLI without re-explaining architecture or…
-
4.2
FareedKhan-dev/kimi-k3-in-c
A portable C99 inference engine that runs the 2.78-trillion-parameter Kimi K3 model on a single CPU within 8.24 GB of RAM, with no BLAS, framework, or GPU. The model streams from disk, and memory only affects speed.
-
4.2
General-Instinct/InstinctFlash
A high-performance serving runtime for robotics models (VLA, WAM, policies) with FP8 acceleration and optimized kernels on Jetson Thor and RTX 4090/5090.
-
4.2
ryana/jevify
A set of prompts for coding agents that helps investigate what TypeSafe's structured decision model Jev could enable in an existing project.
-
4.1
debpalash/VoiceStudio
A fully local, open-source ElevenLabs alternative for voice cloning, voice design, video dubbing, dictation, transcription and audiobook creation on your own hardware, with no account or API key. Supports 16 TTS and 11…
-
4.1
OnlistTeam/ai-manager
A local-first desktop app for installing, updating, and configuring AI coding CLI tools (Claude Code, Codex and others), and managing their API endpoints, Skills, MCP connections, and prompts without hand-editing…
-
4.1
simonlin1212/TradingAgents-astock
A fork of TradingAgents tailored for China's A-share market: 7 AI analysts (including policy, hot money and lockup watchers) run bull/bear debates, risk assessment and produce investment conclusions on A-share data.
-
4.1
tt-a1i/archify
An agent skill for Cursor, Claude Code, Codex CLI, and OpenCode that turns a system description or codebase into interactive HTML/SVG architecture, workflow, sequence, data-flow, and lifecycle diagrams from typed JSON…