tzafrir / human-in-the-loop

A Claude Code mod: Claude assigns you the tasks only you can do. They wait in a pane until you answer, and your answer goes back to Claude.

LingyeSoul / DSH-SoulMojito

把 DSH（deepseek-harness）扩展成兼容 SillyTavern 生态的角色扮演工作区。角色卡、世界书、预设、MVU变量结算、Mod 扩展——资产与数据都在你自己的机器上，LLM 跑在你自己的 DSH 路由与原生 AgentLoop 里

jackispm / nausicaa-harness

Nausicaa treats an Agent run as a dynamic topology of Lanes rather than one fixed linear loop. The model can decide when to observe, fan out, delegate, or collaborate across Runs; the Host keeps identity, permissions, durability, and recovery as explicit facts.

kielchang / dooping-design-book

—

pungme / superagent-desktop

The desktop home for your coding agent — Claude Code or Codex. A persistent chat, a real browser it drives on your own logins, your phone in the loop, and routines on a schedule. Local, on your own subscription, open source. macOS.

muffin-project / muffin-agent

Muffin — one owner-run personal agent: continuity of memory, work and authority across models, tools and surfaces.

TypeSafeAI / jev-harness

A custom coding harness for TypeSafe AI's Jev: an LLM proposes, Jev answers narrow questions, code decides, every step leaves a receipt.

until-dev / plugins

The outer loop for coding agents. Define intent in a Plan, let the agent run, and check every pull request against it.

Sy-D / handraise

Human escalation for Solari browser agents — hand off the live session, solve the blocker, resume in the same session.

MichengAI / dsh-pua

DSH PUA — 为 DeepSeek Harness 提供 PUA 任务推进、角色风味、验收循环和运行状态卡片 · PUA task persistence, personas, verification loops, and live status cards for DeepSeek Harness

AZHi-xinxin / haiguitang-coop

人机共猜海龟汤——人类与AI轮流提问、推理、评分、揭底的双人协作游戏。纯语言、强互动、支持自定义题库。

navels / neal

A plan-driven, multi-agent coding loop: separate planner, coder, and reviewer roles, each on the provider and model you choose, work together to implement your plan.

l1veIn / rambledesk

Generative UI for working with coding agents. Review, annotate, brainstorm, and give feedback through task-specific desktop workbenches.

yinhong-zhou / jevdo

Just Jev it. A Jev-first agent loop with reusable actions for DeepSeek Harness.