📚 An educational, step-by-step implementation of LLM KV cache reuse — learn prefix caching, chunk keys, storage & eviction, from pure Python to GPU tensors and vLLM integration. Inspired by LMCache.

System One Coding Agents: code-level answers at the same score as Claude Fable 5.1 on the same 74-point xz benchmark (69 / 74, 93.2 % each), at 1/20 of the cost and 3.5× faster. No Context Loss and Memory Limits on sessions since every answer starts from the same, updatable, SQLite index.

TestOPS一站式AI运维平台 · By 一鸣（公众号：一鸣AI编程）

确定性取证 × 大模型定性 × Agent 反证复核的恶意代码审计智能体 · Deterministic forensics × LLM verdict × agent falsification for malicious-code audits

This repo is a alternative solution for standard self-attention. It uses tree routing with learnable functions for cheaper self-attention while not compromising accuracy.

Qwen3.8-Flash-Next (180B MoE) on one 12-16 GB NVIDIA GPU, Linux or Windows. Own engine, own compression.

MCP server that lets AI agents use your Firefox, Zen, or LibreWolf browser.

Turn any local language model into a fast decision API with typed, probabilistic answers

Make AI agents aim higher without losing control: a Claude Code skill with two bars per task, a cheap gate check before expensive work, and an independent verifier that never sees the worker's reasoning. Plain markdown, EN + RU.

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

Curated list of open-source LLMOps tools: tracing, evals, AI gateways, guardrails and prompt management. Sorted by stars, refreshed weekly.

A decision library for running local LLMs on multi-GPU machines without NVLink: parallelism, communication, quantization, speculative decoding, KV cache. Every conclusion carries its evidence grade and reproducible conditions.

Five Claude Code plugins, Python stdlib + git. ar-board: judge agent proposals one by one on a local board; the agent keeps your reasons. publish-check: scan a repo's history for personal data before it goes public. tidy-board: delete stale branches, worktrees, and stashes only on approval. Plus doc, handoff, and TODO tools, and a status line.

An experiment-driven LLM engineering lab with 9 reproducible Python experiments exploring token economics, API costs, caching, context growth, latency, model routing, rate limits, concurrency, and batching using Groq.

An experiment-driven laboratory for understanding LLM sampling strategies. Explore how Temperature, Top-k, and Top-p influence token probabilities and generation diversity through reproducible simulations, visualizations, and real Groq API experiments.

Repositorio oficial de skills para el entorno Mexicano caótico.

data agent 从问数到数据开发治理，自然语言问数、SQL 查询、图表看板与多 Agent 专家团分析，支持 MCP 数据源、自带模型和本地 Ollama。

Python voice AI assistant: microphone STT, Google Gemini answers, offline TTS with pyttsx3. Clean env-based setup.

开源 Agent Skill：一键部署 QQ AI 群友（AstrBot + NapCatQQ 全自动安装配置）

Gravitas is a free add-on for Google Antigravity that makes its AI agent check what your answer depends on, tell you kindly when something you assumed is wrong, and still give you what you came for.

A skill for AI coding Agents; Free OEM service manuals -> procedure, torque table, labelled figures. 69 makes, 1960-2025. Agent skill + stdlib navigator.

A reference chart of neural network architectures, from the perceptron to diffusion LLMs, drawn from one spec as a poster and an interactive page. Inspired by the Neural Network Zoo.

人人都能成为 AI 博主：从零起号、AI 产品实测、内容制作、商业合作，59 张操作卡 + 30 天打卡｜AI creator handbook (Chinese)

Research internship cold-email tool: finds professors matching your resume, writes personalised emails, sends them paced from your inbox, tracks replies. Free, open source. By Avinash Chandra Gupta.

Analyze jailbroken iOS and macOS apps, then build and verify independent implementations of selected behavior.