Havuz

The agent that evolves with you 𖤍

STEER — AI4R (AI for Research) experiment framework for auto-research and autonomous research. Connect existing research projects and let an

Asking for What Was Never Requested: horizontal and vertical proactivity in LLM agents. Need-graph metrics (no LLM judge) and Q&D, which tra

Pre-inference tool and agent routing for MCP, A2A and tool-rich LLM systems. Reduce candidate tools before model inference with policy-aware

Local-first Agent RL workbench for Search-R1 rollouts, reward evaluation, GRPO/verl training-data handoff, and an auxiliary private RAG work

27 scope-enforced AI agents that run the full pentest kill-chain (recon → exploit → post-ex → DFIR → report) as a one-command Claude Code pl

A sandbox service for running untrusted, agent-authored code that reports the isolation boundary each run executed behind. Lessons, no clone

SWE-Serve: an agentic benchmark of 53 production inference-engineering tasks derived from merged SGLang pull requests, run with Harbor.

Skill to generate the knowledge vault for projects using the Ralph loop

Let Claude Code, Codex and Pi talk to each other: cross-session messaging that wakes AI agents on one machine or across your LAN (macOS/Linu

Official Findings of EMNLP 2026 implementation of Corpus2Skill: compile a document corpus into a navigable skill hierarchy that LLM agents e

The control plane for your coding agents, however you run them. One checkout of rules, skills, subagents, hooks and stances, projected into

A shared, git-backed wiki your team's AI agents maintain but can't break — Claude Code, Codex, OpenCode and any MCP client.

ALGAL is a programming language for agents. Programs pause for approvals, survive restarts, replay from recorded effects, and improve throug

RHO: Evolving Agents in the Dark — Retrospective Harness Optimization via Self-Preference. Improving LLM agents from unlabeled past trajecto

Always-on 2D lidar SLAM with named places, as a spatial memory for LLM agents

AI agent skills for thinking clearly, making better decisions, and getting things done. Strategy, research, decision-making, planning, risk

SDK, skills, and examples for Inkbox: give AI agents a persistent identity with email, phone number, and internet address

An open taxonomy and scoring framework for evaluating AI agent sandboxes: 7 defense layers, 7 threat categories, 3 evaluation dimensions, 27

Autoresearch agents that improve your benchmark: independent re-measurement, enforced budgets, results as pull requests

Coding Agent Runtime & Evaluation Harness for controlled repository-level software repair

The control layer for agentic AI. A protocol-aware proxy that applies identity, policy, trust and payment controls to A2A, MCP and DIDComm t

Open source skill library for AI coding agents to write, optimize, and debug high performance compute kernels across CUDA, Triton, and quant

给AI伴侣增强时间感的小器官（电脑端专用）

HOP 3.0 — HopSpec task-spec language + HopJIT execution engine for LLM agents

Change management, SkillOpt optimization, and evidence-gated evals for agent skills.

Async Python bot that scrapes YouTube trends, analyzes viral strategies with Gemini LLM, and dispatches HTML reports to Telegram.

A git-native, review-gated knowledge base for AI agents: they propose writes, you approve them. Every claim cites a source, every change is

Reusable AI skills for Codex, Claude Code & Copilot CLI — Markdown-only, safety-first samples for diagnosing bugs, drafting changelogs, and

A minimal coding-agent harness built to measure what context management actually costs: constraint survival after compaction (8%→83%), def

Claude Code plugin for autonomous AI research — multi-agent loops take a bare topic all the way to running experiments, with no human-writte

:star: AI-assisted development workflow that produces understanding alongside code. Teach, capture decisions, document, assemble reviews.

Build & prompt Qwen3.x agents in their native dialect — extracted from qwen-code & Qwen-Agent

Use Codex and Claude Code with COMSOL via sim-cli: .mph inspection, live Desktop sessions, bounded execution, artifacts.

Jetpack Compose & Kotlin AI agent skills for Cursor, Claude Code, Codex, Gemini & 27+ agents. Strict MVI · Kotlin 2.x K2 · Compose 2026 · CI

Can LLM/MARL agents coordinate in long-horizon, open-ended tasks? NeurIPS 2026 (Evaluations and Datasets).

PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents

F.R.I.D.A.Y — a secure self-improving personal AI agent. Local-first, with memory, tools, voice, MCP, automation, and verified self-extensio

Evidence-backed audits, evaluations, and measured improvements for any AI agent.

Self-hosted personal AI agent with a WebUI, memory, tools, MCP, sub-agents and automations. Connect local or cloud LLMs.

242 evaluated academic Claude/agent skills across 18 research domains — bioinformatics, cheminformatics, data science, databases, clinical,

Benchmarking data agents for verifiable analytics over heterogeneous workspaces.

An open-source, coding agent harness designed specifically for local LLMs. Designed as everything a user needs to jumpstart their own harnes

Human as Agent（人即智能体）· Human-as-LLM 人工代理网关：把工程师变成模型，OpenAI 兼容 /v1 接入 Agent 调度池，涉密/需人工任务路由给真实工程师。为 AI Agent 提效、为平台/运维收集工单。Human-as-LLM Gate

Your coding agents forget. Graft doesn't. Persistent local memory that brings back fixes, decisions and project knowledge when they matter —

Writ: An AI agent skill that catches AI writing tells and clichés using a self-edit checklist drawn from Wikipedia's Signs of AI Writing gui

Gates 84.7% of tool-schema prompt tokens away for DeepSeek Harness: plugins declare themselves organs, a nerve impulse routes each command,

IvyClaw, yazılım geliştirme görevlerini birden fazla AI agent, araç ve sandbox kullanarak uçtan uca yürütmeyi hedefleyen Python tabanlı bir sistemdir.

Neden yükseliyor · Repo 15 günlükken 61 yıldız almış; bu, kısa sürede ilgi gördüğünü gösteriyor. Ancak son 28 günde 4 commit ve 1 farklı yazar var. İlginin kaynağına veya benimsemeye dair ek kanıt yok; yıldızlar tek başına üretim olgunluğunu doğrulamıyor.

Agentic Runtime

面向"面试官 × 简历主人公"场景的个人简历 RAG 问答 Agent：把简历、证书、荣誉、项目介绍放进知识库，面试官可对简历任意提问（自我介绍、项目深挖、技术八股、HR 素质、简历细节），系统以第一人称自然作答。核心设计是"宽松推理 + 诚实标注"——硬事实（数字/日期/证书名