编辑甄选 · OSS

GitHub 高星项目· 3 projects

由潜龙编辑团队甄选，每周更新最值得看的开源项目。我们看重的不是 Star 多少，而是它解决问题的方式。

全部主题高潜项目 agent ai autonomous-agents cli coding-assistant context-memory developer-tools knowledge-management llm mcp memory typescript agents chatgpt function-calling gpt gpt-4 gpt4 information-retrieval language-model llama llm-agent ai-agents ai-companion digital-human digital-life jarvis-assistant lip-sync streaming talking-avatar talking-head voice-agent voice-assistant webrtc agentic-ai ai-tools anthropic claude claude-code code-sandbox container-security containers incus ai-assistant ai-cybersecurity ai-hacking ai-security-tool blackbox-testing ctf-tools knowledge-graph mcp-tools penetration-testing ai-sdk chatbot gemini gneerative-ai multimodal openai react agentic-workflow antrophic codex codex-cli gpt-5 indie-hacker macos ai-agent-tools cronjob-scheduler data-science database llms open-source orchestrator ai-agent assistant automation bridge codex-app-server slash-commands wechat celery deepagents deepseek docker faiss-vector-database flask langchain langgraph llmops accessibility android-automation gemma4 litert local-llm on-device-ai phone-agent pocketclaw pokeclaw tool-calling ai-coding coding-agents llm-inference ollama openrouter agentic bedrock genai litellm machine-learning android emulator ios mobile physical real simulator artificial-intelligence cncf hacktoberfest istio k8s knative kserve kubeflow kubernetes ai-agent-memory ai-chat ai-context-extraction ai-context-management alfresco arcadedb docling document-processing falkordb graphrag hybrid-search agent-identity agent-infrastructure agent-memory agent-orchestration agent-state ai-memory bun local-first azure-ai chatgpt-api claude-3 claude-ai copilot copilot-chat gpt4all groq a2a agentframework agentic-framework agenticai aiagentframework aiframework agents-sdk ai-agents-automation claude-cowork generative-ai multiagent opeani ai-native-language ainl compiler deterministic-execution domain-specific-language dsl graph-ir langchain-alternative llm-orchestration agent-auth agent-authentication agent-indentity agent-scaling ai-backend aiagent cloud-native agent-swarm code-review coding-agent github guardrails agent-harness dingtalk-robot feishu-bot llm-wiki mcp-protocol multi-agent openclaw plan-and-execute skills ai-integration cursor game-development gamedev model-context-protocol control-plane distributed-agent-runtime governance mcp-security observability claude-skills gemini-client grok langchain4j ai-security bug-bounty ctf cybersecurity exploit exploit-chaining hacking-tools nmap offensive-security api-server command-line-tool gguf huggingface huggingface-models huggingface-transformers inference-server llamacpp local-ai clawdbot clawhub enterprise-ai gdpr hermes hermes-agent agent-framework agent-skills antigravity cursor-ai ai-gateway anthropic-api cli-tool cliproxyapi developer-productivity github-copilot multi-account nodejs oauth multi-agent-systems personal-assistant python security self-hosted telegram-bot-ai-assistant computer-control computer-use computer-use-agent full-stack gemini-ai domain-adaptation enterprise enterprise-solutions obsidian-md obsidian-plugin personal-knowledge-management pkm aider autonomous ci-cd cline devops github-action github-issues ai-agents-framework chatbots framework javascript llm-observability agent-evaluation ai-agent-benchmark benchmark browser-agent browser-automation browser-use chrome-agent chrome-extension dataset ai-os ai-skills aiden aiden-agent electron moltbot open-source-project bi-temporal-graph code-intelligence code-search agent-os agent-sandbox agentic-rag agentos application rag ai-coding-assistant byok claude-code-alternative codex-alternative cursor-alternative desktop-agent desktop-application ai-cheap cost-optimization cost-tracking gateway llm-router personal-ai self-improving langchain-python event-driven fastapi low-code monitoring nextjs openapi orchestration pydantic claude-code-plugins marketplace cursor-agent desktop-app git-worktree ai-team gemini-cli icor logseq-alternative markdown agi ai-suite ai-workspace deepseek-api gemini-api librechat multi-modal multi-model fine-tuning on-prem swe-bench vscode ai-developer-tools context-engineering amd gpu local-server mcp-server mistral npu agent-teams agentic-coding claudecode code-quality kvcache-compression kvcache-optimization kvcache-reuse lmstudio macbook mlx mlxllm mlxstudio omlx matrix gateway-services llm-cost llm-gateway load-balancing mcp-client ai-gateway-support envoy envoyproxy llm-proxy llm-routing anti-sycophancy cognitive-architecture hebbian-learning hipaa mind-palace ai-presentation api gamma powerpoint-automation powerpoint-free powerpoint-generation presentation

共 3 个项目·第 1 / 共 1

排序Star 最多最新发布

连

CoplayDev

连接大模型与 Unity 编辑器的桥梁

9.9k

简单说，这个项目想把大模型的能力直接接入到游戏开发引擎中。Unity MCP 利用模型上下文协议（MCP），在 AI 助手（如 Claude、Cursor 等）和 Unity 编辑器之间建立了一座桥梁。通过这个工具，开发者可以让 LLM 直接获取操作 Unity 的能力。比如，你可以通过自然语言让...

C#高潜项目1.1k

0xSteph

Pentest-AI：攻防安全 MCP 服务器

515

这是一个将网络安全工具与大模型结合的硬核项目。Pentest-AI 提供了一个专注于攻防安全的 MCP 服务器，里面封装了 205 种安全工具，配备了 17 个专家智能体，甚至还有针对 OWASP Top 10 漏洞的探测器。这里有个小看点：它不仅支持命令行界面，还深度集成了 MCP 协议，允许你...

Python高潜项目97

带

dcostenco

带有记忆宫殿的认知架构

137

这个项目的方向非常学术且硬核。它试图为 AI Agent 构建一个符合 HIPAA 标准的安全认知架构，被称为“记忆宫殿”。项目结合了本地运行的 LLM、赫布学习机制、ACT-R 激活扩散模型以及持久化记忆。它不像是在做一个普通的工具调用框架，而是试图在本地设备上模拟更复杂的多智能体心智模型和认知...

TypeScript高潜项目24