Free local guardrail for AI coding agents: every tool_call judged on-device (regex + Laya), allow/review/block

🎯 The ultimate OSINT toolkit: investigation graph, confidence scoring, alias families, map, timeline and reports. 49 modules, 400+ sites, no API keys required.

小食谱：自托管菜谱库(28 万条) + 吃什么转盘 + 按餐记的饮食记录，AI 拿着 food 工具搜菜谱、教做菜、帮记饭、给饮食卡打分 | Self-hosted recipe book, dinner wheel & meal log with an LLM tool

Production & research-grade Post-Quantum KEM Double Ratchet protocol (FIPS 203 ML-KEM-768 + FIPS 204 ML-DSA-65). Resilient against Harvest-Now-Decrypt-Later FTQC attacks

Self-hosted Discord-style chat and voice for friends. Voice rooms, screen share, push-to-talk. FastAPI + SQLite + LiveKit, ~120 MB RAM, one-command install.

Governed Postgres access for AI assistants: an MCP gateway with OAuth 2.1 and OIDC login, one Postgres role per person, database-enforced read-only queries, PII masking, human-approved writes, a tamper-evident audit log, a red-team suite and Terraform for Cloud Run.

Autonomous, local-first multi-agent AutoML platform. Zero cloud costs, zero API keys, 100% private. Benchmarks 7 algorithms, tunes hyperparameters, and serves live in-browser predictions.

Local Android vision agent: ESP32 BLE/USB HID input, ADB screenshots, a browser workbench and verifiable actions.

RAG knowledge-base chat app built on LangChain — FAISS vector search, FastAPI backend, Next.js WebUI, OpenAI/Gemini compatible

AI 运维监控平台 · AIOps monitoring platform: alerting, AI diagnosis, guard-railed recovery, log forensics — AGPL-3.0

JavSP fork — Web UI + Windows desktop exe + Docker/NAS deploy; 渠道健康熔断 + 本地LLM翻译 + 应用内一键升级; 中文界面 · GPL-3.0

DedSecLab — open-source cyber lab: a Windows PyQt6 client, OpenVPN-gated FastAPI backend, and ForceCommand SSH ops menus in one tree. Run it on your own VPS with your own profiles.

基于 LangGraph 的 AI 旅行规划：对话补全需求、地图与交通查询、约束优化、可视化行程和持久化任务。

AI 模拟面试官：每条扣分都强制引用你回答的逐字原文（机器校验，编造引用会被重试/丢弃），四维度加权由确定性代码计算而非模型现编，评分体系设计为可经真人盲评校准 · FastAPI + Vue3 + LangGraph

Intelligent Document Investigator & Cross-Document Contradiction Engine

Edge-cloud vision reference: TFLite gate, MQTT, LLM fallback, temporal memory, calibrated metrology, and a beginner ESP32/serial kit

AI 驱动的量化分析平台（开源版）：因子分析 / 选股 / 多因子轮动回测 / 模型实验室 / 自动因子挖掘 · Qlib + Alpha158 + FastAPI + Vue3

Built for Smart India Hackathon 2026, SIH Problem Statement 26192 by National Disaster Response Force (NDRF), DM Division, Ministry of Home Affairs: Hyperlocal flash flood prediction and risk-response system for hilly regions using multi-source environmental and IoT data.

tsov (the shape of voice) — hum it, get a song you can keep editing. A local-first, AI-native music workspace: humming → transcription → arrangement → mixing → export, with humans and agents sharing one command layer.

Tipos e validadores de dados brasileiros de alta performance para Pydantic v2 e FastAPI (CPF, CNPJ 2026, CEP, PIX, Telefone, CNH, PIS, SUS, Título e mais)

Google Calendar-style weekly view of your Claude Code sessions — local, live, read-only

Real-time offer recommendation with Bayesian Thompson Sampling. SQL feature pipelines, a FastAPI service on Docker and AWS with a sub-100 ms target, automated retraining, offline evaluation and backtesting against random, greedy and UCB, and monitoring for regret, allocation, latency and drift.

Full-stack AI short-drama pipeline with FastAPI, Pydantic structured outputs, ComfyUI and FFmpeg

A production-grade reference implementation for LLM agent systems: flow declared as a LangGraph graph, Pydantic AI for typed agents, generated WebSocket and REST contracts, human approval on irreversible actions, tracing and evals.

Scaffold production-ready apps with Next.js, TanStack Start, or React + Vite and a separate Hono or FastAPI API. Choose modules for auth, storage, cache, email, payments, jobs, observability, and deployment; add, remove, or upgrade them with three-way merges that preserve your edits. Each CLI release pins a tested Clubedge Starter.

Agents propose. People approve. Change control for AI agents: everyday work runs on its own, anything an agent can't take back waits for a teammate's sign-off, every action on a tamper-evident record. Open source.

A local support-agent recovery lab with Ollama, durable tool execution, independent outcome verification, and reproducible failure-scenario evaluations.

📉 Revenue resilience × price-drawdown backtesting engine — five-panel research workspace, 3D CAGR surface sweeps and out-of-sample factor search on SEC EDGAR + Yahoo Finance data.

3DGS 在线重建查看器 + 个人主页（郑州大学双足实验室 3D 方向考核作品）

Salt & Story turns voice memos of a relative cooking into clean, printable recipes, while keeping the speaker's own words beside them. Record or upload a memo, an open-weight model (Gemma 4) writes it up, and you export a cookbook PDF that looks like a gift. It is built for one real family: Because Every Recipe Has A Story