Pensieve: never lose the thread. A local MCP server and Mac app that links your files, repos, agent sessions, tickets and messages into one thread per task, for you and your AI coding agents.
Open and Reproducible Benchmarking for Relational Learning
Minimax video audio Concat AV Latent for any video into Minimax H3 sampler
Value-guided action selection and self-improvement for frozen behaviour-cloning policies
Meet JevEmbed — turn embeddings into decisions. Choose, score, and judge with your choice of embedding model.
Context-Matched Distillation: Teacher Causality for Autoregressive Video Distillation
Long video in, scored vertical clips out — the clipper that shows its work. Local-first OpusClip alternative: speaker-tracked camera, styled captions, auditable virality scoring, Instagram outcome calibration.
SRC-6K Security Research Skills: black-box SRC and white-box audit agent package
Enables tongue tracking and eye convergence on a rooted Quest Pro
Datasets, scripts and results for the paper "The Pain Axis" by Valen Tagliabue, Leonard Dung & Cameron Berg
Distill a repeated Jev classification task into a local model, on the fly — same answers, your hardware.
RynnValue: Scaling Robotic Value Foundation Models with Temporal Distance
—
MariaDB 13.0.1-rc RCE lab — priv-esc + heap UAF + JOP chain to system() as uid 999(mysql) on stock Docker image. Found with RAPTOR and raptor-loop-hunt.
Qwen 3.8 27B at 48-65 tok/s on a Mac Studio M4 Max — measured oMLX native-MTP recipe, same-day A/B vs mlx-vlm+MTP sidecar, raw benchmark JSON included
A nightly newspaper that writes itself: an agent runs the desks, the figures come from code, and it refuses to publish an edition with marks in it.
PanoVLN: panoramic vision-and-language navigation research code
挖兔硬盘精灵 —— 免费无广告的 Windows 硬盘健康守护:SATA/NVMe 检测 · 六档颜色预警 · 开机静默体检 · 一键 HTML 报告 · 单文件绿色版
BACnet-style priority arbitration for Home Assistant: give every command a level of authority, so a manual action and an automation can disagree without one silently clobbering the other.
Code review via the change impact measure
CTF多层内网渗透辅助工具
—
关于gpt的全自动/半自动注册机,可以自动导入到gpt2api使用,能够配合gpt2api实现自动补号。
Wave Forcing: Towards Speed-of-Light Streaming Video Generation
Stealing Reasoning Traces from Proprietary LLM APIs
Ontology-driven e-commerce analytics system / 本体驱动的电商数据智能分析系统
—
Explore sleep, training, mood and nutrition together using local records, traceable calculations and optional MCP tools for your preferred AI client.
OpenJev: open-source, local alternative to Jev. System-One decision models: typed questions in, calibrated probabilities out; general decisions and browser-agent steps.
AI API gateway / relay: 307 models behind one OpenAI-compatible key. image2.5 $0.0085/image, Seedance 2.0 Mini $0.01056/sec, LLM from $0.0228/M tokens. USD settlement, pay-as-you-go, $1 minimum top-up.
一个端到端的全自动流水线项目,覆盖 ChatGPT/OpenAI 免费账号注册的完整链路:用你自己的 iCloud 隐藏邮件地址(Hide My Email)生成 @icloud.com 别名作为邮箱源,自动注册 GPT 账号并通过接码平台完成手机验证,最后生成 Codex OAuth 凭证并导入 VPS/代理。仅供自动化你自有资源的账号注册与授权,每个账号独立邮箱 + 独立手机号(一号一账号),不绕过平台验证机制。
DeepStack: Facilitating Co-Design Exploration of 3D DRAM-Stacked Accelerators for Distributed LLM Inference. Includes the MICRO 2026 AE artifact.
Lightweight self-hosted web search & extract service for Hermes Agent, a drop-in Firecrawl alternative
A local tool for verifying references in research papers. It extracts citations from PDF and docx files, searches for matching scholarly records online, and creates an HTML report for human review.
Build fonts for Xteink e-readers running CrossPoint, and tune them with live preview
BackSearch plugin for Hermes Agent — point-in-time web search & fetch over a frozen news archive (General Reasoning)
An open-source JEV-powered workbench for streaming data selection, quality evaluation, automatic LoRA training and held-out model evaluation.
—
Agent Skill: turn a character image into a looping transparent sticker pack (WebP/GIF/PNG + ZIP).
A Python/Pandas ETL pipeline that turns a deliberately messy automotive dataset into production-ready data, with 15 pytest tests and GitHub Actions CI.
Standalone Robinhood Chain LP explorer, accounting and research terminal
A pure-Python reachability and skeletonization layer that decides what a coding agent actually sees before a request reaches the API.
把预算 Excel 变成可拖拽、可交互、数字精确的「预算沙盘」网页 · Turn a budget Excel into an accurate interactive cash-runway sandbox. Fixed verified engine — AI never rewrites the math.
羽化面板 - 基于 sillyGirl 的 AutMan 兼容聊天机器人面板
A concise Chinese portrait prompt builder for Z-Image in ComfyUI
Android security and penetration testing, from first principles to senior practitioner. 81 chapters, ~327 labs, measured rather than recited.
墨枢是一个面向中文长篇网文作者的 AI 小说写作工作台,支持作品大纲、章纲、人物与世界观设定库、时间线、RAG 检索、一致性守卫、自然化审查、AI 辅助续写、团队协作、用量计费,以及面向漫剧改编的静态分镜管理。前端采用 Vue 3,后端采用 FastAPI,支持 PostgreSQL/pgvector、Redis、Docker 自部署和 Cloudflare Workers 预览。
Ultra-tiny 2.1k streaming voice activity detector (2.1 KB INT8). 200 ms window for microcontrollers, edge devices, and DSPs. ships with ONNX, TorchScript, and standalone C header.
—
Discover agent rules and skill files from leading open source projects
Data, code, figures, and supplementary materials for a study of 142 Arabic- and English-language fact-checked misinformation claims from the 2026 Lebanon conflict, comparing authentic-content misuse with AI-related content and examining newsroom penetration.
逆向 FLUX Kontext Max API: 1K $0.064. model ID `flux-kontext-max`, OpenAI-compatible, async task + cost in every response, $1 minimum top-up. flux-kontext-ext-reverse-api-cn.
Group3r but in Python
Open-source paper-trading research app integrating TypeSafe Jev, animated 3D model pets, and a safe offline demo.
Patcher for Cyber Troopers Virtual-On (PC, 1997). Makes the game run nicely on modern hardware - with online support to boot!
Open-source, model-agnostic AI company research for long-term investors. · 面向长期投资者的开源、模型无关 AI 公司研究系统。
Persian Speech to Text benchmark
—
A third-party desktop novel-writing tool compatible with DeepSeek Harness.
—