The OpenAI Agents SDK (Python) is a production-grade library for building agentic AI applications. It provides a complete agent runtime (`Runner`, `Agent`, `RunConfig`) with support for tool calling, handoffs, guardrails, sandboxed execution (Docker, Modal, E2B, Cloudflare, Vercel, Unix-local), realtime voice/audio pipelines, MCP integration, and multi-provider model routing (OpenAI, LiteLLM, AnyLLM, custom). The codebase is async-first with extensive tracing/observability, session persistence (SQLite, OpenAI Conversations, Responses Compaction), and a rich lifecycle-hook system. It includes a large test suite covering unit, integration, and sandbox security scenarios, plus internal tooling skills for release management and code review. The SDK is actively maintained with a 0.23.x version, release automation via release-please, and comprehensive docs in English and Japanese.
Ruflo is a large, modular AI agent orchestration platform that extends Claude Code with a plugin ecosystem, swarm coordination, memory systems, neural learning, security enforcement, and federation. The core is a CLI tool (`npx ruflo`) that provides 26+ commands and 215+ MCP tools for managing agents, memory, security, hooks, and deployments. It includes a console UI inside Claude Code, a Svelte-based chat UI (ruvocal), and a React-based goal UI. The architecture is ADR-driven with clean separation into domain, application, and infrastructure layers across 35+ plugins and 20+ core packages. The system implements advanced patterns: CQRS for memory and swarm, circuit breakers for resilience, capability envelopes for security, Ed25519 signing for release integrity, and a harness loop for automated champion selection. It supports multi-model routing, hybrid vector search (HNSW + SQLite), watermarking (WASM-based Gumbel scheme), and neural learning (PPO, DQN, Decision Transformer). Federation enables cross-installation agent collaboration with zero-trust, HMAC-signed envelopes, and PII-gated data flow. While the codebase is well-structured and actively maintained, it has some maintenance risks: duplicate helper scripts across multiple directories, hardcoded version numbers in smoke tests, experimental alpha plugins, and a large regression test suite that requires ongoing upkeep. The plugin ecosystem is extensive but some plugins may be aspirational rather than production-ready.
This repository implements an AI-powered code review CLI tool (`ocr`) that analyzes Git diffs or full file contents using configurable Large Language Models (LLMs) such as Anthropic, OpenAI, and AWS Bedrock. It originates from Alibaba Group's internal code review assistant, validated at massive scale (tens of thousands of developers, millions of code defects). The core workflow involves loading a diff or file set, dispatching review tasks to an LLM agent with tool-use capabilities (file reading, code search, comment posting), and collecting structured review comments with line-level precision. It supports multiple review modes (workspace changes, single commit, branch range), full-file scanning, session persistence with resume capability, token budget management, and output in text, JSON, and SARIF formats. The architecture is a Go CLI with a Cobra command tree, complemented by a VS Code extension (TypeScript/Preact), plugins for Claude Code, OpenCode, and other coding agents, and CI/CD integration examples for GitHub Actions, GitLab, Bitbucket, Gerrit, GitFlic, and Alibaba CodeUp. A local web viewer (`ocr viewer`) and a public documentation/marketing website (React) are also part of the repository. The codebase exhibits strong engineering practices: extensive test coverage (unit, integration, manual e2e), clean package separation, thread-safe components, and careful error handling (retry observers, host guard for the viewer). Key reusable patterns include a retry collector for LLM API calls, a state-machine-based run manifest builder for coverage tracking, an abstract file reader for workspace vs git-ref reads, and a worker pool with per-key async processing.
DeerFlow is an open-source, full-stack AI agent harness built on LangGraph/LangChain, providing a production-ready platform for building, deploying, and managing conversational AI agents. The system features a Python/FastAPI backend with a modular architecture that includes agent orchestration, long-term memory (DeerMem), sandboxed code execution, a skill system with static analysis (SkillScan), MCP (Model Context Protocol) tool integration, multi-channel IM support (Telegram, Slack, Discord, Lark, etc.), scheduled tasks, subagent batches, and a comprehensive authorization framework. The frontend is a Next.js/React workspace with real-time streaming, artifact management, and a terminal UI (TUI) for headless operation. The architecture follows a gateway pattern: the FastAPI gateway handles authentication (JWT, OAuth, PATs), routing, and SSE streaming, while the core harness package (`deerflow`) implements agent logic, middleware chains, and runtime components. Persistence uses SQLAlchemy with Alembic migrations, supporting both SQLite and PostgreSQL. The system is highly extensible via a Python extension API (`deerflow-extension-api`) and a plugin system for frontend surfaces. Deployment supports Docker Compose and Kubernetes (Helm chart). While primarily an application, DeerFlow includes research-oriented components such as memory eviction benchmarks, context snapshot evaluations, and tracing integrations (LangSmith, Langfuse, OpenTelemetry). The codebase demonstrates strong engineering practices with extensive unit tests, blocking IO regression tests, and contract tests, though its large feature surface and many configuration options introduce complexity and potential for drift between environments.
AgentMemory is a persistent memory system designed for AI coding agents (Claude Code, Cursor, Codex, etc.) that captures agent observations via hooks, compresses and stores them, and enables semantic/recall search across sessions. Built on the iii-engine framework, it provides a modular architecture with CLI, REST API, MCP server, and web viewer interfaces. The system eliminates repetitive context re-explanation by automatically injecting relevant memories into agent sessions, implements memory consolidation to extract stable facts, and includes features like working memory management, audit trails, and health monitoring. Agent integrations occur through platform-specific hooks and MCP tools, allowing agents to store observations and retrieve context without manual intervention.
CowAgent is a production-grade multi-agent AI assistant platform that routes messages from 12+ IM channels (WeChat, Feishu, Telegram, DingTalk, Slack, Discord, QQ, WeCom, web console, desktop Electron app) through a modular agent core. The core implements planning, tool use (bash, file editing, browser, web search, MCP, sub-agent delegation, scheduler, memory, vision), three-tier memory with vector embeddings and markdown-aware chunking, knowledge base management, skill system with SKILL.md manifests and a skill hub, self-evolution (idle-conversation review loop), and multi-agent team orchestration with peer transport. The architecture is a modular monolith: channel adapters normalize messages into a unified ChatMessage, the Bridge dispatches to per-agent/session model resolution, the Agent protocol (Task/AgentResult/AgentAction) drives the execution loop with streaming, cancellation, and steering, and all state persists to SQLite with incremental migration. Key entry points are app.py (startup orchestration), cli/cli.py (daemon lifecycle), channel/web/web_channel.py (HTTP API and SSE streaming), and the desktop Electron client (React/TypeScript with Zustand stores). External dependencies include LLM APIs (OpenAI, Claude, Gemini, DeepSeek, etc.), IM platform webhooks/SDKs, optional voice TTS/ASR providers, and MCP servers. The codebase demonstrates strong engineering practices with extensive regression tests (200+ test files), atomic file writes, thread-safe context propagation via ContextVar, and security guards for credential paths and SSRF.
This repository is a meta-skill/plugin for Claude Code that serves as a 'team-architecture factory'. It is not a traditional software application but a collection of highly structured Markdown files (SKILL.md and references) that act as instructions for the Claude Code AI to design and evolve multi-agent teams. The system takes a domain description from a user and generates agent definitions, skills, and orchestrators to automate complex workflows. It demonstrates deep expertise in multi-agent orchestration: three execution modes (workflow orchestration, persistent agent collaboration, sub-agent delegation), six team architecture patterns, and workflow-native quality patterns (adversarial verification, judge panels, loop-until-dry). The project underwent a ground-up rebuild from v1 to v2 to align with the current Claude Code runtime, removing reliance on deprecated APIs like TeamCreate. The codebase is exceptionally well-documented for its domain, with clear principles, phased workflows, testing methodologies, and migration guides. The primary artifact is documentation, not executable code, which limits traditional software engineering evaluation but the craftsmanship in structuring meta-instructions for an AI is high. It is actively maintained with recent commits, thorough changelog, and professional issue templates.
This repository is an educational course on harness engineering for AI coding agents. It is a collection of lecture materials, code examples, and project templates that teach how to build environments, state management, verification, and control mechanisms for AI coding agents. The course is structured as 14 lectures and 8 projects, with a sample Electron knowledge-base app used as a running example. The repository is primarily documentation and educational content, with some reusable templates and scripts for harness creation and auditing.