Havuz

Claude Code usage governor: compact professional output, context slimming, tool-output filtering, telemetry, and drift guardrails.

Ten cooperating AI agents that turn long podcasts into short-form videos, automatically. Free and open source, runs on free AI providers.

A CLAUDE.md built on Karpathy's principles and gstack, adding a verification harness and autonomous-loop rules for Claude Code.

AI-powered SIP PCAP analyzer with call flow, hold/resume detection, and GPT-4o chat

Automatically completes the full workflow from requirement research → research review → planning → plan review → development → development

Learn what AI skills are and how to design, structure, and use them in real-world agent systems.

the agi compiler: records llm agent behavior, proves what repeats, and compiles it into verified, sandboxed wasm binaries that run for micro

Your second brain - capture your mind

Self-hosted, Docker-first OpenAI-compatible LLM gateway with provider routing, fallbacks, API keys, and usage controls.

A drop-in proxy that compresses bloated code context in real-time, cutting LLM API costs by 50–80% without losing what the model actually ne

An open-source text language, parser, validator and Three.js renderer for inspectable 3D human movement.

Alignment-research scaffold (autoresearch-style) for LLM guardrails: search over a single policy.md surface

An AI writing skill that turns agent-generated drafts into clearer, warmer, more human documents by removing templated phrasing, markdown cl

让Codex向你呈递公文并接受批示圈阅

Cross-platform Claude Code desktop client by hwwn. Multi-workspace, remote mobile access, session branching, BYO LLM. macOS / Windows / Linu

MemPrivacy is a privacy-preserving personalized memory management framework for edge-cloud agents.

Production LLM call layer for AI agents and tools: keep OpenAI/Anthropic/AI SDK/LiteLLM, hot-swap models with MDA presets, and add cache, re

Evaluation Metrics for African Languages

GPT-2-style LLM built from scratch in C/CUDA with hand-written backprop, BPE tokenizer, FlashAttention, pretraining, and SFT.

让 AI 主动记错题、自我迭代的协作机制 · A framework that makes your AI reflect on its own mistakes and stop repeating them

A benchmark for evaluating LLM × harness performance.

现编OS — 一台不预装任何软件的电脑，每个应用都是 AI 当场现编的 macOS 风格网页OS · An OS where every app is improvised by AI, live

An open-source cost controller for AI agent loops — stops a loop when it's actually converged and rolls back before it degrades, instead of

An engineering philosophy that treats the agent as a worker: the system prompt is a JD, skills are certifications, model tiers are degrees,

Memoriki - LLM Wiki + MemPalace. Personal knowledge base with real memory.

原生AI股票研究平台，支持多智能体分析、AI Agent 协作与实时辩论对抗，可配置的大语言模型（LLM）工作流、多源市场数据及报告导出功能。

Automate viral reel/tik-tok videos via AI multi-agent system. ~$0.1/reel.

LoRA fine-tune Gemma 4 31B to speak caveman-mode natively. Style: github.com/JuliusBrussee/caveman

The marketplace for verifiable AI outcomes: state a task, get a fixed price upfront, and Primitive Bench routes across tools to complete it,

Karpathy's understanding ladder as agent skills: auto-pick text, diagram, interactive HTML, or video storyboard for any explanation. Nothing

Self-hosted AI SRE for Kubernetes — zero-instrumentation eBPF observability plus a copilot that fixes issues through guardrailed, self-verif

Agent Deck

🔍 Reverse engineer any codebase into a build prompt — architecture breakdown, ASCII blueprint, and AI-ready reconstruction prompt

CLI Tool for 8k context ai models

Alephant is an open-source AI Agent Gateway for routing, tracking, and controlling LLM usage across AI agents, members, and workflows, and f

Governed, provenance-sealed creative rendering for Databricks — a Streamlit App + Unity Catalog ai_render() function that turns any governed

⚡ Cut 60-80% token cost for DeepSeek V4 Pro on Codex CLI. token-saver + memory skills with prefix-cache optimization.

一个让 Claude Code 调用 Codex 干活，并可以同时调用多个模型（GPT、Kimi、DeepSeek 等）的 MCP 工具。

Source code of paper "RLCSD: Reinforcement Learning with Contrastive On-Policy Self-Distillation"

Cheaper and better Greptile alternative runs on your own github actions.

Anyone can predict the future. Almost nobody keeps the receipt. A hash-chained prediction ledger for LLMs — sealed before the answer exists,

A native Python agent CLI built on DeepAgents CLI, featuring an independent memory Agent that captures learnings after each task and deliver

GBase — Recursive Self-Improvement Agent Framework. Memory, evolution, quality gates, identity system, and 40+ auto-registered tools.

QuantClaw is a plug-and-play task-type routing quantization plugin for OpenClaw.

Drop-in AI memory layer with 2x faster retrieval and 10x lower cost. Fully compatible with Mem0 API. Migrate in 5 minutes without any code c

Dataflow-Oriented Reinforcement Learning for (Multi-)Agentic LLMs

Drive your real Chrome from any AI agent — logged-in sessions, real fingerprint, live tabs. CLI-based, works with OpenCode, Pi, or any Harne

AI-powered CLI for reviewing GitHub pull requests, detecting potential bugs, security risks, regressions, and missing tests, with structured

Distributed 56M-parameter LLM inference across 3 ESP32-S3 boards via ESP-NOW , Split-PLE + KV cache, fully offline.

🦀 Decoder-only LLM built from scratch in pure Rust using Candle — no Python, no PyTorch. Gated DeltaNet + sparse attention, fine-grained MoE