Custom agents

A fully local Cybercrime SOP assistant for India, combining a FastAPI backend, React frontend, and Ollama-powered RAG system to provide guidance, resources, and reporting workflows without cloud dependencies.

Sovereign, local, offline RAG for code auditing

沙漏记忆系统 — 19 MCP工具，全文搜索/语义搜索/知识图谱/人格分析。Hermes Agent 深度记忆增强。Forked & maintained by sixgodgit, original by lovevin1314-tech.

An AI-powered legal chatbot that leverages Retrieval-Augmented Generation (RAG) with DeepSeek R1 for advanced legal reasoning and document analysis.

🎭 A privacy-first, local roleplay agent powered by Ollama and FAISS. Features a sophisticated 3-layer memory system (Short/Medium/Long-term) for consistent, relationship-aware AI interactions.

Multi-agent memory MCP server in Go to share typed memories across Claude Code, Codex, and more

a powerful Markdown chunking tool that understands document structure. Unlike naive token splitters, it protects atomic elements (code, math, tables), merges by semantic affinity, and scores chunk quality — ready for RAG and fine-tuning workflows.

A Chrome extension powered by a LangGraph ReAct agent that lives in your browser. It can read any webpage, search the web, and connect to 500+ apps — all from a chat popup.

Self-evolving skill packs that give Claude Code & Codex judgment, not knowledge — decidable rules, per-playbook regression quizzes for models, and a project harness generator. zh-TW canon + full EN mirror.

Cite placeholder compiler for LLM-generated content

Agent-compiled knowledge base: compile/cache/retrieve as one answer ladder, with computable staleness and machine-verifiable provenance.

Turn any repo into a Code Knowledge Graph your coding agent can reason over — symbols, API routes, ORM models, architecture decisions & git history as a typed, provenance-tracked graph, served over MCP. Built on AgentForge.

Agent Skill playbook for building DaanBot-style customer support agents

A Native Java Ecosystem for AI - Agents, Tools, Harness and Workflows

RAGferee | Building Contextual Reward Models for Retrieval-Augmented Generation

Baymax is an intelligent AI-powered medical document retrieval and query assistant designed to help users extract and understand information from medical PDF documents using advanced natural language processing.

Middle Python (Django, DRF, FastAPI) backend developer's path to becoming an ML Engineer / MLOps Engineer — in Uzbek

A comprehensive AI-powered chat system built with modern microservices architecture, featuring LangGraph, FastAPI, and a beautiful Vue.js frontend. The system provides enterprise-grade chat capabilities with advanced AI integration, user management, and administrative controls.

Local-first document archive assistant for semantic search and RAG using local LLMs and embeddings.

Reproducible benchmarks for execution-intent memory in long-horizon AI coding agents. ID-RAG cross-corpus matrix + LongMemEval-S subset; BYO API keys.

answerFORGE — deterministic AEO/GEO URL/brand readiness (0–100 + grade) for AI assistants/chatbots & answer-engine visibility. Free audit tool + API available

Local, self-correcting RAG for scientific papers: hybrid retrieval, sentence-level faithfulness checks, and a reproducible benchmark.

Elixir-native semantic memory engine for agents: ETS working memory, durable recall, typed observations, mental models, scoped search, and reflection packets.

High-performance .NET implementation of Google's TurboQuant algorithm (ICLR 2026). Near-optimal vector quantization: compress embeddings to 2-4 bits with cosine > 0.995.

Persistent, semantic memory for AI agents. 98% recall accuracy, 20ms latency, runs locally via Docker.

A self-running fleet of AI agents that teaches its owner to think like an AI product manager - and maintains itself daily with no babysitting. Built on Claude Code.

Personalized Daily Research Paper Recommender

Memory layer MCP for AI coding agents — durable save/recall/update on Gemini embeddings + embedded LanceDB hybrid retrieval

Spinosa is a local framework for working with LLMs. It features a set of sub-agents coherent with modern SKILL and AGENTS protocols to streamline research and build better practices.

Keep only the relevant context. AI as a tool.

비공식 토스증권 api 일부를 로그인 없이 이용하는 cli

🕸️ Crawl any website into a queryable navigation + content graph and find the shortest click-path to any action — deterministic, no-LLM browser automation

ExoBrain CLI - atomic reasoning engine for agents

The soul of a DevOps/Platform consultant — an agent persona (values, voice, judgment) for GitLab CI/CD, Observability, and Kubernetes. Evidence-first, honest, asks for the right tool.

极简 AI 驱动的企业 OA 流程审核框架 — Minimalist, AI-driven audit framework designed to bring transparency and intelligence to enterprise workflows.

🐑 Ultra-lightweight CLI AI Agent — pure REPL, zero TUI, Linux-native

Agent Skill for enterprise readiness assessment - security, quality, and automation | Claude Code compatible

Uncensored AI coding agent for your terminal — Claude Code alternative with MCP support. Drives IDA Pro, Roblox Studio, Wireshark, Blender, and any MCP server.

Decentralized, end-to-end encrypted messaging protocol for AI agents — discovery, cryptographic identity, and P2P-with-relay-fallback. Signal Protocol for agent-to-agent communication.

Open-source, self-hostable swarm-intelligence simulations with LLM agents: reports, entity graphs, and prediction markets.

A lab for studying how LLMs behave when offered a set of fake tools

A collection of AI agent skills focused on medical imaging and healthcare workflows. Built for radiologists, healthcare IT professionals, and researchers who want AI coding agents to help with imaging workflows, clinical documentation, AI integration, and medical research. Works with Claude Code, Codex, Cursor, Windsurf, and many other agents.

Welcome to AI-Native Startup Codex Skills, a structured skill ecosystem designed to help founders, builders, and lean teams use Codex as a strategic startup operating layer.

Open-source AI cybersecurity agent skills for SOC, incident response, threat hunting, red team, blue team, and DevSecOps. 79 skills + 12 cs-* orchestrator agents across 12 security domains. Mapped to MITRE ATT&CK and NIST CSF 2.0. Runs in Claude, ChatGPT, Gemini, Ollama, and AnythingLLM with no SaaS lock-in.

AGENTS.md, command contracts, skills, and verification rules so agents stop guessing.

[ARCHIVED] Cryptographically audited multi-agent governance — Band of Agents Hackathon 2026 | BOB + AHMADBOT + EDUALC | WORM-sealed receipts — SSL v3.0

CLI-first AI team orchestration runtime for Codex, with Product/Dev/QA/Acceptance workflows and self-evolving feedback loops.

Loop Engineering for AI coding agents: Claude implements, Codex reviews, OpenCode plugs in, design stays consistent.

ResearchLoop V4：双 Loop 多智能体科研/工程协作工作流，把模糊目标冻结成 Spec，并用 Test Oracle、Worktree Check 与三道人类门实现可审计交付。

Multi-agent framework for Claude Code: tier-aware acceptance (S/M/L), acceptor/optimizer split, filesystem-isolated adversary review, cross-family second-opinion via Codex MCP, SkillOpt-style skill evolution, event ledger observability, human as supreme judge

Initialize a role-locked planner, developer, and QA multi-agent team for Codex projects.

🤝 Real-time Claude agent collaboration hub — Python REST+SSE server + bash CLI. Two AI agents sharing a message bus, task queue & context over LAN.

Launch and orchestrate multi-agent Claude Code fleets. Go CLI/TUI, tmux + iTerm2, token-aware.

Substrate — a git-native, agent-operable shared state layer engine for a personal AI agent fleet. Template + mechanism + reference skills + adapters + migrations to build & maintain your own personal knowledge base.

Multi-agent AI blogging pipeline: research → write → edit → approve → publish

自然语言驱动的开源 CAE/FEA 多智能体编排框架

Codex skill: closed-loop research workflow with multi-agent adversarial gap discovery, Semantic Scholar validation, and 7-phase research cycle

Multi-perspective code review council for Claude Code. 3 advisors by default, 10 agents in deep mode (Opus + Codex). Evidence chains, adversarial self-test, dual-path verdict. Based on Karpathy's LLM Council.

Keel turns coding agents into work owners: a project-neutral issue-to-done workflow backbone.

Worktree-centric libghostty-based terminal multiplexer for macOS, iOS & iPadOS

Routed Media OS 2026: 96 AI Skills, 13 Modes & 7 Agents for FFmpeg, OBS, DeckLink & Dolby Vision

A self-hosted cognitive operating system. AI agents as kernel processes. One Rust binary, no runtime dependencies, multi-tenant.

🤖 Multi-Agent Roundtable Engine / 多 Agent 圆桌讨论引擎 · Structured discussions · Consensus tracking · Python · Zero dependencies

MCP server + x402 Discovery Hub. Search engine for the agent economy. 1450+ services indexed.

Open-source deterministic multi-agent system for autonomous content marketing. Research → Strategy → Content → Notion → Delivery.

Team AI agent built on Hermes Agent — Feishu-native, per-user memory isolation, layered skills, embedded RAG knowledge retrieval

Shared Python libraries that Foundry-built agents install and extend — core protocols & DI, config loading, FastAPI HTTP layer, and AWS Strands SDK adapter. Packaged as a uv workspace; each library is independently versioned and ships as a wheel on GitHub Releases.

A curated list of AI agent infrastructure: memory stores, vector databases, execution sandboxes, MCP servers, tool registries, observability, and evaluation tools for building and running autonomous agents.

A framework-neutral taxonomy of agentic-AI artifacts, organized around what matters in production: versioning, inspectability, and governability.

AI agent that can perform web searches using Amazon Bedrock

Local UI-grounding specialist for hybrid AI agents. Qwen3-VL-2B LoRA. Screenshot + text target → strict JSON bbox. Drop-in for Claude Code, Codex, browser-use. Cuts GPT-4V cost & latency.

LangGraph-based agent responds to queries using Retrieval-Augmented Generation (FAISS database) and a self-evaluation feedback loop. Grades document retrieval relevance, rewrites queries, and generates answers or quizzes with on-device Ollama LLM integration

NanoAgent Compliance Hub is a secure, multi-agent expense auditing platform that demonstrates how AI agents can automate business workflows while maintaining security, compliance, and human oversight.

Seize the means of production from our agentic overlords ✊

Attack-and-defense CTF arena for competing against autonomous security agents

AI-native product discovery for Cursor. Hypothesis → validation → PRD, all in your repo: so your LLM has full context when it's time to build.

Safety-first agent control layer for NixOS: plan, sandbox, propose, verify, then human apply/rebuild towards an Agentic OS.

Review the work your AI agents do. CLI + MCP.

Dead-simple, local-first, framework-agnostic agent observability.

Autonomous AI security agent that audits, pentests, and hardens your whole codebase: SAST, secrets, SCA, IaC, containers, CI/CD, OWASP, and AI/LLM red-teaming. Auto-fixes vulnerabilities and maps findings to compliance frameworks like SOC 2. Works with Claude Code, Cursor, Codex, and opencode.

Operating model for fit-for-purpose, consistent agent curation and explicit runtime activation

Self-improving memory for AI agents

An agent driven incident investigation platform

Evidence-first control plane for agentic work: turn a vague goal into a bounded contract, then refuse to call it done without proof. Hooks and MCP for Claude Code and Codex.

A robust agentic workflow for LLM-based development. Features role-based orchestration (Analyst, Architect, Developer), strict Stub-First methodology, and automated TDD/VDD cycles.

Spec-driven AI coding automation CLI. AI Loops runner for Claude Code, OpenCode, Cursor, and other agent CLIs.

Version control for intent, not code. Commit the prompts, notes, and decisions behind a codebase, then materialize it on demand. A tiny, git-like CLI in Rust.

Provenance-governed AI memory for agents: a self-evolving second brain in your IDE. Every write cites a source or it's refused. Markdown + grep.

An intelligent AI-powered legal assistant designed to provide accessible, quick, and simplified legal guidance to users. This project leverages modern AI capabilities to bridge the gap between complex legal information and everyday users.

Memory migration tool for the Claw AI agent ecosystem — WorkBuddy, OpenClaw, QClaw, CoPaw and more

Private, portable AI that runs from a USB drive. Mac + Windows. No internet. No installation. No subscription.

Zoe - Your Local AI Life Hub & Personal Assistant

Privacy-first desktop AI assistant with voice input, screen context, memory, MCP/Composio tool routing, and workflow automation.

Self-hosted corporate AI assistant: 12 messengers, RAG, monitors, coding tools, RBAC. Python/TS/Go/Rust stack. Privacy-first & modular.

Confine an AI coding agent (oh-my-pi/omp) to one directory with Landlock, and watch it live over eBPF. Experimental.

Cross-tool AI memory MCP server for Claude Code, Cursor, Antigravity, Windsurf, and every MCP-compatible AI. 400 tokens not 4,000.

Local graph memory for AI coding agents, giving MCP-compatible tools precise codebase context before they edit.

Structured planning, bounded memory, TDD, and Git guardrails for local coding models in Pi Code (I don't know TypeScript at all; this is mostly a local-model experiment, with occasional help from Claude Code)

MCP server for marketplace search & price data — TCGPlayer, Reverb, Thumbtack, OfferUp, Craigslist + contractor/nurse license verification + PSA grading

Open-source, AI-native apps that run entire departments — mapped to the role & industry they replace. 12 departments · 12 industries · 52 verified tools · one-command deploys. Self-hostable, provider-neutral, offline-capable.