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 ai ai-agent ai-agents connector dag fastapi function-calling llm mcp nextjs openai-compatible python agent-observability analytics claude-code cli codex coding-agent cost-tracking dashboard developer-tools devtools agent-builder agent-orchestration agent-workspace agentic-ai ai-automation ai-coworker ai-skills-manager ai-workflow-builder harness agentic-bi agentic-workflow agents ai-workflows artificial-intelligence business-analytics business-intelligence langchain langgraph rag automotive car doip edge-ai home-assistant llama-cpp local-llm obd offline-ai agentic-coding anthropic claude claude-code-alternative cpp agent investment langraph skills trading acp agentic-development-environment code-review coding-agents github-copilot golang kanban agent-security ai-agent-security ai-antivirus ai-security cursor devsecops dsh-plugin gemini-cli chatbot crm customer-support ecommerce lgpd multi-tenant nuvemshop open-source chatgpt hermes hermes-agent openai data-pipeline data-processing machine-learning sdk typescript vscode-extension calendar finance gdrive gmail google-calendar notion productivity scheduler shopify ai-coding ast cline code-intelligence context-window copilot cli-tool codex-cli developer-productivity hooks linux llm-tools agent-os ai-coding-agent deepseek deepseek-harness dsh agent-skills claude-plugin comfyui comfyui-extension comfyui-mcp comfyui-mcp-server flux image-generation local-first ai-tools browser-ai browser-automation browser-extension chrome-ai chrome-devtools chrome-extension chrome-extensions firefox-addon firefox-addons agent-memory ai-infrastructure knowledge-graph llm-memory mcp-server memory-management multi-agent-systems agent-management ai-assistant ai-worker cross-platform desktop-app digital-persona model-management openclaw agent-skill claude-code-plugin claude-code-skills claude-skills coding-assistant empryo graph docker knowlege-graph n8n neo4j ollama pgvector ci-cd clickhouse evals evaluation llm-as-judge llm-evaluation llm-observability llm-ops ai-sdk decentralized-ai llama llm-inference local-ai local-fine-tuning nodejs on-device-ai open-source-ai open-source-llm open-source-rag agentic-framework ai-chat automation autonomous-agents gemini llm-agents mcp-client ai-coding-assistant aider bubble-tea discord ai-content ai-marketing awesome-claude content-creation content-engineering linkedin 近半年高星 agent-team fimga flutter opencode openpencil accessibility android autonomous kotlin multimodal-ai call-graph-analysis code-indexing code-search developer-tools-ai-agent mcp-tools cybersecurity holehe maigret model-context-protocol osint rust terminal tui ai-development java reactjs agile-workflows claude-ai code-analysis codebase-audit documentation-generator agile-scrum apple doubao minimax parallel-agents api-gateway continue goose kilo oh-my-pi openai-api chatgpt-alternative dify flowise homelab make-alternative electron apis data-flow development low-code no-code browser browseros chromium hacktoberfest lmstudio macos windows agpl ai-agent-platform audit-trail odoo rbac self-hosted telegram-bot autonomous-pentesting bug-bounty ethical-hacking penetration-testing pentest pentesting-tools recon gguf agent-tools ai-workflow command-line-tool go mcts-rl meta-agents meta-optimization runtime-supervision tree-rl workflow-automation demo easy embedding flask gpt huggingface-transformers multimodal qwen code coder extension ide vscode cost-reduction llmops mitm-proxy prompt ai-platform enterprise clap emby jellyfin jellyfin-plugin k3s kubernetes librosa lyrion music navidrome agpl-v3 ai-safety auditing blockchain dcap-attestation decentralized-marketplace hardhat intel-sgx polygon-edge api embedded-ipaas integrations ipaas anthropic-api apple-silicon deepseek-v4 diffusion inference code-understanding codebase-search graph-database bolt cypher database enterprise-solutions graph-rag graphql hnsw memoryos mvcc instant-messaging

共 3 个项目·第 1 / 共 1

排序Star 最多最新发布

团

Agenta-AI

团队级 AI Agent 构建空间

4.6k

随着团队内部开始大量尝试 AI Agent，如何协同开发和管理成了一个新问题。Agenta 定位为一个供团队构建 Agent 和自动化流程的工作空间。它不像是在做一个单一的聊天机器人，而是提供了一套基础设施，涵盖了 Agent 的编排、技能管理以及可观测性。真正难的地方在于如何让非硬核开发者也能参...

TypeScript高潜项目640

Jwuthri

Tracely-ai：专为 AI 智能体设计的 CI/CD 测试工具

965

真正难的地方不是写一个能跑通的 AI 智能体，而是如何保证它在修改提示词或更新模型后不会突然崩溃。Tracely-ai 试图把传统的 CI/CD 理念引入到 LLM 开发中，提供了一种基于追踪（Trace）的回归测试方案。它的工作流很巧妙：自动捕获生产环境中的失败案例，将它们聚类并冻结成独立的测试...

Python高潜项目35

编

mikehasa

编程 Agent 行为与成本追踪

663

随着大家开始频繁使用各种写代码的 Agent，一个新问题出现了：这帮 AI 到底在后台干了什么，又花了多少 API 费用？这个项目就是为了解决这个痛点。它提供了一个本地优先的仪表盘，专门用来追踪 Claude Code、Codex 等编程助手的行为。它会把每个任务拆解成具体的步骤，包括使用了什么工...

Python高潜项目74