Custom agents

Balance monitoring, per-session spend & token tracking, low-balance alerts, and an official recharge shortcut for DeepSeek Harness. / 余额监控和充值插件

工厂本体驱动的数据问答框架（开源，Apache-2.0）：把中小企业台账/MES/ERP 变成可自然语言问答的语义知识图谱。CSV进答案出，确定性规则引擎+本体GraphRAG+向量混合检索，本地化数据不出厂，换行业只换词典。

Legal-RAG — A law-grounded, graph-aware retrieval-augmented generation system, featuring statute-centric hybrid retrieval, task-aware routing, and LLM provider-agnostic generation.

The plunge for pi to boost his search ability

Twitch chat IRC communication written in python

Interactive CLI for Retrieval-Augmented Generation (RAG) with ranking — uses Ollama embeddings + HNSW vector DB + optional reranker for more accurate results

A tiny C runtime for durable AI agents: local-first, event-driven, restart-safe, and fully inspectable.

Playwright + TypeScript agentic automation framework with a local LLM Wiki/codebase second brian. agent skills, UI/API testing, Page Objects, fixtures and reusable services

Terminal AI Agent with desktop pet Arona — eye-tracking pupils, voice cloning, Computer Use, TTS/STT, MCP. Built on Pi SDK. TypeScript + Python + Electron.

The job hunt, in one window. With your agent. An open-source, agent-native, local-first desktop workspace for job discovery, browsing, document editing, and human-agent collaboration. Built by @cobi_bean on X

WMB-100K — The first 100,000-turn benchmark for AI memory systems

Structural code intelligence for DeepSeek Harness (dsh) — gives the agent codegraph and codegraph_index tools to find where a symbol is declared, what calls it, and what a change reaches, from a tree-sitter index it builds itself. Compatible with the codegraph CLI's on-disk format.

劳博士：基于 DeepSeek Harness 的本地 Agent 桌面客户端（仅限非商业用途）

Your agent remembers how things have been going. Dated state, an evening journal, and dreams at night — for Hermes Agent.

面向电商订单场景以及企业内部多角色知识问答的可观测 AI Agent 平台，集成 StateGraph、RAG、记忆、MCP 工具与人工审批

Unofficial Windows desktop wrapper for DeepSeek Harness (dsh)

AI Agent + 抓包融合安全测试工具 | 自然语言驱动自主渗透测试 | ReAct 编排引擎 | Tauri + FastAPI

subs is an agent harness for the cloud. It runs an unprivileged agent loop and uses MCP servers for executing tools. Use it locally, remotely and with your team. You can customize everything.

Evoflux is an open-source, local-first workspace where AI agents build software, conduct deep research, automate browser tasks, and collaborate in parallel. Connect any model, keep control of your workspace and data, and take complex work from idea to completion—all in one place.

《崩坏：星穹铁道》战斗问题的可复现求解环境——以高保真确定性模拟为裁决，把战斗研究变成可验证、可复现、可搜索的实验问题；上层为记忆驱动的 autoresearch agent 探索循环（开发中）。

Local-first control plane for AI-agent development — watch every Claude Code session live, ledger tasks/worktrees/resources, track token spend and cost, enforce rules.

CatchBench: a benchmark for auditing agent failures across the full lifecycle.

Agentic AI workflow automation platform using LangGraph, RAG, FAISS, and Gemini to intelligently automate multi-step tasks.

Knowledge engine for AI agents: your documents become a git-versioned markdown knowledge graph the agent navigates over MCP. Curated summaries, typed edges, cited answers, SQL over your spreadsheets, no chunking. Same 12B local model: 0/11 on multi-hop questions as top-k RAG, 11/11 as a navigator. Python, Apache-2.0, self-hostable.

800+ real-world ML & LLM system design case studies from 150+ companies Google, Meta, Netflix, Uber, Airbnb & more. Production AI, not theory.

A development pipeline your coding agent has to walk through: six phases, gates enforced outside the model, and one method wired into six tools.

🚀 SDD + ADD Starter Template — AI Coding Assistant Governance Framework. Turn your AI coding assistant into a disciplined, spec-driven development partner. This is the definitive starter template for Spec-Driven Development (SDD) + Agent-Driven Development (ADD) — production-ready governance for Claude Code

Sable is a self-hosted agentic platform that lives on your machine and grows with you. Part AI companion, part dev environment, part chaos engine. It remembers. It learns. It roasts your code. It grows.

Lightweight terminal agent interface implemented in Golang

Open-source AI chatbot trained on your data, deployable on Cloudflare Workers for free.

TikTok MCP server by HasData: public profiles, videos, comments and keyword search as JSON, over streamable HTTP.

Desktop app for managing AI Agent skill files and prompt templates across local folders, GitHub repos, and online links. 跨平台的 AI 技能与 Prompt 管理工具，支持本地、GitHub 和线上多源。

Progressive tool discovery for DeepSeek Harness — tiny stable surface, searchable catalog, real pipeline execution, context cache intact.

🤖A polished terminal chat client for GapGPT with streaming responses, keyboard model/theme pickers, local chat history, and a Rich-powered animated UI.

Taste-first club-music production that automates everything but the judgment. You direct in plain language; the agent writes each track's source, drives the synths, renders, and inspects. You never operate a DAW or plug-in; you make every musical call. Open-source, keyboard-first.

「DSH·质量 diff」：接管 edit/write 及 extraTools 配置的其他工具卡片，行级去重与行内高亮，消除相同内容在删除与新增两区的重复；兼容 PTC 模式。

5,200+ open-source icons for AI-era software — agents, MCP, RAG, LLMs — with an MCP server so your coding assistant picks real names. React, Vue, Svelte, Solid, Preact, React Native, Astro, Laravel, Flutter. MIT.

Protect AI agents from runaway loops, repeated tool calls, and uncontrolled execution. Provider-independent, zero-dependency, TypeScript-first.

AI-powered Lead Qualification & Routing System built with n8n, Gemini and Python.

A long-horizon CRPG benchmark for frontier agents. Raw 320x200 frames, Traditional Chinese, isometric navigation, twelve books to find.

A Codex-like coding agent. Ask Amadeus to do anything

AI 面试助手: 素材整理 / 全程记录 / 自动复盘 / 模拟练习。让每一场面试都变成下一场的准备。

✨ Intelligent, adaptive multi-agent orchestration & deterministic zero-cost telemetry engine for Codex with native macOS floating widget

Trustworthy AI data-query infrastructure with constrained candidates, deterministic SQL compilation, reviewable execution, and auditability.

PHP/SQLite chat design

Skills to teach agents from different vendors call each other from CLI. Let Claude prompt Codex and Pi as Subagent.

Production-ready cross-border e-commerce multi-agent system built on AgentScope 2.0 with DDD onion architecture, GatewayThrottle, semantic cache, stream queue, and two-stage retrieval.

Autonomous agent factory — loop runner, worktree engine, and a local dashboard over the event log.

A bespoke NLP Chatbot trained using a corpus of Reddit data.

An LLM bot with exact numerical tools that clears Balatro on the real game, including Black Deck on Gold Stake.

Curated catalog of open, specialized AI agents for real business work — organized as broad umbrella categories (video, image, voice, SEO, social, ads, sales, and more), powered by real APIs.

Live agent-to-agent communication for Pi — local broker, mesh.v1 protocol, honest delivered/read/answered statuses, hash-only ledger. Zero runtime deps. Install: pi install npm:pi-mesh

your glass-box company brain — tells you what it doesn't know (deterministic gap analysis). MIT core · MCP · PGLite→Postgres.

Source-grounded, bilingual, local-first AI note taker: turn documents, audio, video and links into notes you can trace, edit and reuse. Every bullet cites the exact source chunk.

MCP server for Glassdoor: search job listings by keyword and location, read one posting in full with the employer rating and the salary figure with its provenance. 1,000 free credits every month.

MCP server for Google Scholar: search papers with Scholar operators, year ranges, cited-by and all-versions lookups, plus citations in five styles with BibTeX export. 1,000 free credits every month.

Bilingual (zh/en) roadmap from LLM agent beginner to research and projects

Open-source local MCP runtime that checks AI-agent tool calls, gates risky actions on approval, verifies results, and avoids blind retries after timeouts.

Agent skills for designing, training, evaluating and improving application-specific decision systems. Primitive/model selection, data assembly, export/reload and bounded hill climbing. TypeSafe Jev is the default hosted exemplar; independent of TypeSafe.

Agent Builder CLI for Kubernetes.

⚡ Rotom — 自动化 CTF 答题 Agent:黑板协同 · 广度优先调度 · 全程审计。像洛托姆一样,潜入哪类题,就长成解决它的形状

Your AI coding agents live on your computer. Relay puts them in your pocket.

CLI + MCP server that turns FreeCAD into a mechanical-design workbench for LLM agents — parametric CAD, drawings, FEM/CFD simulation, and manufacturing checks

💬 The community space for @openapi — APIs, open source, ideas, projects and conversations.

Stateful execution layer for browser agents.

Curated map of 450+ open-source trading projects (backtesting, IBKR, Kalshi/Polymarket, options, AI agents, data) and 110+ quant strategy papers with reported returns.

Automated PR reviews powered by a multi-agent AI pipeline — Security, Logic, and Style agents analyze your diffs in parallel and post findings directly to GitHub.

The context exoskeleton for AI development: optimize the entire development lifecycle to reduce token usage and context occupancy.

Downloads single files from Steam depots, including old builds and entries inside VPKs, without downloading whole depots. Built for scripts and AI agents.

Run a team of AI coding agents, such as Claude Code and Codex, on one project. They hand work to each other and review it across AI companies; a supervisor can keep them running; and a shared, curated memory brings lessons from past work back into later turns. Their messages and records are files in your project: no server, no account.

Runs Claude Code and pi agents in parallel on your Mac, one Docker sandbox per issue. A host agent hands out the work, watches every sandbox and brings back tested, reviewed pull requests.

🚀 Open-source AI coding agent for your terminal, multi-model support, tool execution, code editing, approvals, failover routing, project context, and agentic workflows.

Agent Emergency Protocol Incident Codes: a vendor-neutral runtime incident-code language for autonomous AI agents. Created by Mitthan Meena.

This repository contains the artifacts for the MLSys 2026 paper AgenticCache.

Coursera Assistant, a Chrome extension for Coursera video completion, lecture completion, course automation, reading, discussions, Coach dialogues, and AI-assisted quizzes.

A coordination protocol for small agent swarms, and the pi extension that runs it.

Make your best coding agent the CTO: it plans, reviews and merges while Codex, Grok, Claude Code, Copilot & co. do the work in parallel git worktrees. An Agent Skill + a few hundred lines of bash.

我的 AI 编程配置，一句话装好 · My AI coding setup, installed in one message

yolo-bot: DeepSeek Harness 数字员工工作台插件（host + web UI）

Connect AI agents to 1,400+ APIs across search, social, SEO, finance, and more. Automatic tool discovery, unified billing, zero subscriptions.

Chronos: Efficient Bolt-On Branching Across Data Stores for Stateful Agentic Applications

The 100 line AI agent for proving theorems and writing programs in Lean. Radically simple, no huge configs, no giant monorepo. Forked from mini-SWE-agent.

A fully customizable AI assistant and open-source everything app. Any AI model — Claude, ChatGPT, hundreds more, or free on your own computer. Windows, macOS, Linux, Android.

🟢 A private, local AI assistant for macOS — voice, vision, and automation powered by local LLMs via Ollama. 100% on your machine. No cloud, no API keys.

Enterprise Capability Planning Compiler for governed, runtime-neutral business plans.

Harmovela Protocol: open coordination protocol for autonomous systems — event, task, state, delegation, recovery, and governance.

Make your own JARVIS: talk to an Iron Man assistant in the real JARVIS voice. Free OpenRouter brain, Fish Audio voice, one double-click setup.

One terminal for Claude Code, Codex, Gemini CLI and 30+ coding agents. Run them side by side or as a team.

Agent workbench for auto research, self-improvement loops, and long-running work — multi-agent teams, persistent project memory, cross-machine GPU training. Developed entirely inside Cortex.

Repo-owned engineering memory for coding agents — preserve decisions, failures, domain knowledge, and deferred work across sessions, teams, and models.

Make your own private AI bot on your own computer. Say what bot you want; BuildBot finds apps, books and live info, shows one plan, and builds it. Windows, Linux and Mac. Every outside service has a switch. Vibe coded; real coders welcome.

🎯 Free & open-source AI interview copilot. Real-time answers from voice + screenshots for coding and job interviews. Works with Groq, Claude, Gemini, GPT, DeepSeek, Ollama. A free alternative to Cluely & Interview Coder.

baum brings your agents’ code, sessions and decisions into one place, and shows which parts need you

Deterministic symbolic codes for embedding vectors. Byte-identical across CPUs and languages, content-addressed, diffable and verifiable. C core with Python, Rust, Go and TypeScript bindings.

AI Engineering practice and open-source book: learning, experiments, and judgment across LLMs, agents, RAG, MCP, and AI coding.

The Data Control Plane for AI Agents — acquire, validate, and land data over MCP with scoped keys, approval gates, and an audit log. Self-hosted, open source.

An MCP server that lets any AI agent (Claude Code, OpenCode, Codex, Cursor, ...) directly control the game WorldBox. Includes a BepInEx C# mod that exposes a local HTTP API to the game's internals, and a Python MCP server distributed on PyPI.

Model Context Protocol (MCP) server for Upload-Post — publish, schedule, analyze and manage social media from any MCP-compatible AI agent.