Search

Pipeline-driven personal AI agent: 10 LLM providers, 7 real tools, 6-stage pipeline, React Web UI + CLI. Zero mocks.

Live-updating index of LLM agent frameworks shipping on GitHub, refreshed every 15 minutes

灵智MindPal — 元数据驱动的超级智能体系统。统一管理数据、知识、设备与智能体，支持可控工具调用、策略驱动执行、完整审计追溯与Skill插件扩展，覆盖从对话编排、工作流自动化到具身智能的全链路能力。安全第一，万物皆可建模。

AI Agent 面试知识库：覆盖 LLM、Prompt、RAG、MCP、Tool Use、Agent 架构、Multi-Agent、LangGraph、Claude Code、Codex CLI、工程化评估、安全与源码解析。

Engineering infrastructure for opencode AI agents: memory, skills, subagent dispatch.

🧠 Five-Layer Memory System v4.0.0 - A universal memory skill for AI agents

OpenSkald is a self-evolving knowledge-native publishing agent.

A tool for literature snowball sampling

🪼 A personal AI assistant that lives in Telegram — every topic is its own isolated Claude Code agent with layered persistent memory and proactive cron heartbeats. Open skeleton of a real, running system.

🎥 Enhance video consistency with comfyUI-LongLook, ensuring smooth motion and prompt accuracy for 81+ frame generations in Wan 2.2.

🎥 Transform your video editing with ComfyUI nodes for UniVideo, enabling seamless video understanding, generation, and editing tasks.

An AI-powered EEG classifier that detects emotional states during music listening and curates Spotify playlists grounded in what your brain actually did.

Server-side, diff-synced mirror of a host's live hierarchy — a host-agnostic pattern for building MCP servers over apps with large mutable trees (Photoshop, Figma, Unity, ...).

OpenAI Agents SDK source connector for Proofwright — capture agent runs, tools, handoffs, and approvals as evidence.

Shared foundation library for the Hikigai SDK ecosystem. Provides the core HTTP API client, structured exception hierarchy, and platform constants used by both hikigai-agentsdk and hikigai-appsdk. Not intended for direct use — install hikigai-agentsdk or hikigai-appsdk instead.

LangChain callback adapter for AeroGraph — automatically record LangChain traces.

Provider-agnostic AI agents and skills framework

A modular TypeScript agent runtime with pluggable tools, skills, and LLM backends.

TypeScript/JavaScript SDK for the AI Agent Runtime Monitor — cost guardrails, hallucination detection & audit trail

Let opencode (and any shell) call Claude Code over the Agent Client Protocol (ACP), with resumable named sessions. Ships an ACP-client CLI plus an opencode/Claude skill.