New repos — GH Trending Explorer

多租户 SaaS 智能客服：LangGraph 五技能诊断闭环 + 转人工真实流转（派单/接管/办结/回流）。FastAPI · PG 主从 · Redis · 本地向量化

收藏夹打架了：肌群探索、训练动作收藏与B站视频观点对照

Local CLI that checks your project for accidental secrets, sensitive files, and oversized assets before publishing to GitHub. No uploads. No runtime dependencies.

robot-hint / HINT

HINT: Human-Intent Inception for Long-Horizon Robot Manipulation

Yzw202011 / OmniSpace

本地优先的全栈 AI 创作工作站：AI对话 · AI漫画 · 漫剧创作 · 写作台 · 知识学习五大模块；vLLM / llama.cpp / Transformers 多引擎本地推理，显存智能调度、舱壁隔离自愈、仅供学习使用，欢迎交流指正。（最新进度（除插件系统）受到协议限制，暂时无法提交到仓库）

Vizcexex / ai-video-analysis-system

本地轻量级 AI 视频上传、预览与整段分析系统 | Local lightweight AI video upload, preview and whole-video analysis

i3T4AN / KADATH

Evolutionary multi-agent runtime that breeds, evaluates, and improves autonomous agents across reproducible epochs to converge on optimization of a goal.

B0yko / tern

Local AI search for podcast and video archives on Apple Silicon: Whisper speech, Apple Vision OCR and SigLIP-2 visual search fused into one ranked list, with a trim editor and FCPXML export. FastAPI, vanilla JS, Tauri. Source-visible, all rights reserved.

solari-sdk / solari-cookbook

—

KaiWU5 / Awesome-AI4AI

Updated weekly — until humanity achieves AI4AI. A survey-grounded catalog of 432 papers on long-horizon agents, benchmarks, harness design, and recursive self-improvement.

chenweixin123 / workbooklens

Local-first linting, testing, semantic diffing, and safe repair for Excel workbooks

jklthinking / retinue

Self-hosted coordination for AI workers: per-task graphs, device/model timelines, progress receipts, and source-aware usage. MIT.

LanternW / AutoPHD

—

low-hands / MyCareer

MyCareer Agent——你的全流程秋招/春招求职Agent助手：岗位采集、简历分析与定制、公司调研、投递管理、面试准备和模拟面试

Router0824 / BidPilot-AI

AI 智能投标作战平台：招标文件解析、多 Agent 工作流、技术标生成、咨询中心、商机监控

ybh-ybh / know-where

知归是一个面向个人使用的 AI 知识归档工具。把内容链接（微信公众号、稀土掘金、GitHub、小红书、抖音、B站）发送给飞书机器人，知归会自动完成内容提取、AI 分类与总结，并将结果沉淀到飞书多维表格中。Send WeChat Article / Douyin / Xiaohongshu / Bilibili links to a Feishu (Lark) bot; auto-extract, AI-summarize and archive to Feishu Bitable. Self-hosted personal knowledge base / second brain.

InkMind-AI / Dyn-3D

TempoVista is the official implementation of Dyn-3D, a benchmark and training pipeline for 3D spatial reasoning in vision-language models. The release contains frame-selection tools, SFT and Kinematic-GSPO training code, benchmark evaluation scripts, Dyn-3D annotations, and checkpoints for Qwen3-VL and InternVL3.5.

mrlong0129 / yominpost

Self-hosted AI social media studio: one piece of material in, platform-native posts out. Format router, self-critique, pre-publish risk gate, real OAuth publishing. Web app + CLI + MCP server + agent skill.

QiushenZou / cn-investment-banking-skills

Agent Skills for Chinese IPO sponsor execution: diligence, filing production, pre-submission QC, regulatory inquiries, and filing refresh.

chqwer2 / Disease-Progression-World-Model

—

q734738781 / CatMaster

—

star7code / minimax-h3-chunk-star7

VRAM-efficient MiniMax H3 for ComfyUI: QKV/RoPE/MLP activation chunking, automatic OOM chunk reduction, and CK/SLA/Sol/Hybrid attention for SM75 and SM80+ GPUs.

ii76 / xijian

**隙见**是一款基于真实评论证据的 AI 内容机会发现与选题决策工具。它不是简单做词云、情感分析或标题生成，而是帮助内容运营从评论区中识别真实需求、争议、疑问和场景缺口，并生成有证据、有评分、有内容结构的选题 Brief。核心表达： > 从评论里，看见下一个值得做的选题。

xsyryry / code-reviewer

Code Reviewer

Simreal-AI / hard-puzzle-benchmark

[Public preview] Hard human-written reasoning puzzles: 749 answered items in 15 topics (707 scored core), with protocol and scorer.

dahan6 / whetstones-lab

Whetstones Lab: measuring coevolution between adaptive malware and behavioral defense. De-weaponized research release — genomes, evolution rounds, telemetry, training harness (stubbed primitives), and lab range tooling. Payloads, real primitives, and C2 infrastructure intentionally withheld. Research and defensive use only.

T-Lab / DocDML

—

Nathanielneil / sae

SAE（Standard Academic Editing）Skill

humanpen / humanpen-skill

—

YZhu0225 / must-bench

—

tttboy123 / jlens-evolve

JLens Evolve v3: evidence-centric evolution platform (autonomous Teacher + local Qwen + official SWE-bench native evaluation). Branch codex/v3-evolution-10000 carries transport resilience, Router-key normalization, per-round temperature beam, and baseline fail-soft.

dreamers-laboratory / video-input-to-modern-llm

Give an LLM eyes on a video without blowing the context window: ffmpeg frame sampling at a chosen resolution, token estimates per model family, budget fitting, contact sheets.

Noperi-00 / Panel-Sloper

Panel Sloper — VPN panel (FastAPI/Railway)

obayalhindi / NetAudit-L2

ARP-spoofing based network visibility tool for Kali inux - active traffic redirection and DNS/TLS-SNI cature for a network you own or administer.

Kai26-Han / miniLLM

从 0 训练大模型，单卡训练一个 200M 参数模型的全流程（train a 200M‑parameter model from scratch）

lzlfwow / PatentClaw

论文到中国专利交底书的证据驱动多Agent生成、固定检查表审查与可追溯多轮改稿系统。

jkjkiiiii / mcdinner-mate

一个人吃饭，但不孤单。基于麦当劳 MCP 的语音陪餐陪伴 Skill —— 不只是点餐，是有人陪你吃完这顿饭。

LiXin97 / DN-MOPD

Beyond Teacher Assignment: Domain-Normalized Multi-Teacher On-Policy Distillation (DN-MOPD)

f24sales / litellm-free

F24 SALES free AI model directory, verified LiteLLM routes, import tools and agent-assisted model research.

vyrnsynx / tufty-arm-create

Arm Create conference firmware for the Pimoroni Tufty 2350 only. The with-filesystem image of MicroPython bw-1.28.0-3 restores the firmware plus the badge apps and pictures. Not the stock release from pimoroni/tufty2350.

HeCC928 / shanghai-strategy-lab

```text Interactive A-share strategy research with staged entries, walk-forward validation, trade-level explanations and offline reports. Built with Python, React and FastAPI.

wbsjkya / CrediFusion

—

xuanheng-tech / snapshot-runner

Deterministic, read-only Git evidence for coding agents and automation. Standard-library Python CLI.

g-Ironfox / AAgent

AAgent:用于决策与现实多功能代理的技术大杂烩框架

nirbhay41120003 / EchoMind

Local-first voice memory assistant — capture speech, transcribe live, and ask a private RAG chatbot grounded in what you've actually said. No cloud required.

Kai26-Han / Auro_agent

一个本地可扩展个人 AI agent 工作台（a local-first, extensible personal AI agent workspace）

hbqdcq / Next-Brick

—

omar-thing / omarthing-python

Python client for the unofficial TikTok API: profiles, followers, videos, comments, live data and search

Mapika / decider

A family of System One-style models fine-tuned from Qwen3.5, designed for one-pass typed decisions with calibrated probabilities.

canivel / dynolab

Dyno Lab — an AI safety, alignment and interpretability research workbench for Apple Silicon. Explore activations, probes, SAEs and interventions with local inference, a Python SDK, APIs and MCP.

xvchujin / umat-skill

Abaqus UMAT Codex skills for constitutive modeling, return mapping, damage, state contracts, tangent auditing, and hierarchical validation.

X-Square-Robot / Xplanner

X-Planner: Event-Structured Task Planning for Embodied Intelligence

r1c7 / CluxMate

AI coding agent with one Python core and three front-ends — headless CLI, Textual TUI, and an Electron desktop. Works with any OpenAI-compatible API, with risk-tiered permissions, event-sourced replayable sessions, and a fail-closed OS-level sandbox.