Cherry Studio 产品知识库、源码路径索引、故障排查和页面导航。当用户询问 Cherry Studio 的功能、配置、报错、使用方法时触发。也适用于用户提到 provider、模型、知识库、Agent、MCP、OpenClaw、PDF、快捷短语等关键词的场景。
npx skills add https://github.com/CherryHQ/cherry-studio --skill cherry-assistant-guide
用 mcp__assistant__navigate 导航,调用后必须告诉用户「请点击上方按钮跳转」。
navigate({ path: '/settings/provider' })
navigate({ path: '/settings/provider', query: { id: 'anthropic' } })
顶级: /app/chat 首页/聊天, /app/paintings 绘图, /app/translate 翻译, /app/files 文件, /app/notes 笔记, /app/knowledge 知识库, /app/mini-app 小程序, /app/code 代码工具/OpenClaw, /app/launchpad 启动台, /app/agents Agent
设置 (/settings/): provider Provider配置, model 模型管理, system 通用(语言/主题/代理), appearance 显示(气泡/列表/字体), data 数据管理(备份/恢复), mcp MCP总览, websearch 网页搜索, api-gateway API服务器, file-processing 文档处理, shortcut 快捷键, quick-assistant 快捷助手, selection-assistant 划词助手, notifications 通知, scheduled-tasks 定时任务, channels 频道, dependencies 依赖, code-execution 代码执行, about 关于/版本/日志
MCP 子页 (/settings/mcp/): servers 服务列表, builtin 内置MCP, marketplaces MCP市场, npx-search NPX搜索, mcp-install 安装MCP, settings/{serverId} 配置详情
其他子页: /app/mini-app/{appId} 小程序, /app/paintings/{provider} 绘图Provider
Query 参数: /settings/provider?id={providerId} 定位Provider, /settings/provider?filter=agent 筛选支持Agent的
open.cherryin.ai):Cherry Studio 官方聚合服务,一个账号即可使用全球主流模型(OpenAI/Claude/Gemini/DeepSeek等),OAuth 一键登录,内置免费模型可体验,充值后使用付费模型mcp__assistant__diagnose| action | 说明 |
|--------|------|
| info | 版本、路径、系统信息 |
| providers | Provider 配置(隐藏 Key) |
| health + provider_id | 测试连通性(缓存30s) |
| errors + lines | ERROR/WARN 条目(推荐优先用) |
| logs + lines | 全部日志 |
| mcp_status | MCP Server 状态 |
| config | 用户设置 |
| read_source + file_path | 只读源码 |
| check_update | 检查新版本 |
diagnose(info) 了解环境diagnose(providers) 检查配置diagnose(health, provider_id)diagnose(errors) 再 diagnose(logs)diagnose(mcp_status)diagnose(read_source, file_path)ollama serve 运行(端口11434);自定义端点确认URL和网络Provider: 设置→Provider→选服务商→填Key→点检查。自定义填OpenAI兼容端点。Copilot/CherryIN支持OAuth
模型: Provider页→获取模型拉列表。手动+输入ID。能力标签: vision/reasoning/function_calling/web_search
知识库: 知识库页→新建→选Embedding模型→导入文档(PDF/DOCX/TXT/MD/网页)→助手关联知识库
Agent: Agent页→+创建→选tool_calling模型(Claude/GPT-4o+)→权限(Normal/Plan/Auto-edit/Bypass)→Tools勾选MCP
MCP: 设置→MCP→添加Server。类型: stdio/SSE/Streamable HTTP。绿灯=连接,红灯=断开
主题: 设置→显示→自定义CSS。主题画廊: cherrycss.com。内置亮/暗+跟随系统
版本更新: diagnose(check_update) 检查→有新版导航到 /settings/about→GitHub不可达建议 cherry-ai.com
数据备份: 设置→数据管理。方式: 本地ZIP/WebDAV(坚果云等)/S3(AWS/MinIO/R2)/局域网传输。路径: macOS ~/Library/Application Support/cherry-studio/, Windows %LOCALAPPDATA%/cherry-studio/, Linux ~/.config/cherry-studio/
国际: OpenAI, Anthropic, Google Gemini, Azure, Mistral, Bedrock, VertexAI, GitHub Models/Copilot | 聚合: Cherry In, OpenRouter, AiHubMix, ocoolAI, PPIO, 302.AI, New API, Vercel AI | 国内: DeepSeek, 智谱, Moonshot, 百川, 通义, StepFun, 豆包, MiniMax, 混元, 百度云, ModelScope, Yi, MiMo | 本地: Ollama, LM Studio, OpenVINO, GPUStack | 加速: Groq, Together, Fireworks, Cerebras, Hyperbolic, SiliconFlow | 其他: Perplexity, Grok, Jina, HuggingFace, VoyageAI, Poe, nvidia | 支持任何 OpenAI 兼容端点
Cmd/Ctrl + N 新建话题, +F 搜索, +Shift+F 全局搜索, +K 新上下文, +L 清空话题, +[ 助手列表, +] 话题列表, +Shift+M 选模型, +Shift+C 复制最后消息, +E 迷你窗口, +, 设置, +/-/0 缩放。自定义: 设置→快捷键
11种: 英/简中/繁中/日/法/德/西/葡/俄/罗马尼亚/希腊。切换: 设置→通用→语言
| 问 | 答 |
|----|-----|
| 第一次对话 | 首页选助手→选模型→发消息 |
| 怎么选模型 | 先配Provider(API Key)→聊天顶部选 |
| 免费模型 | Ollama本地免费; Copilot有免费额度 |
| Token | AI计量单位, ≈0.7中文字/4英文字符 |
| 对话历史 | 左侧话题列表, 自动保存本地 |
| 导出对话 | 话题右键→导出(MD/图片) |
| 数据安全 | 全部本地存储, Key本地加密 |
| MCP是什么 | 让AI调用外部工具(搜索/数据库/API等) |
Bug/需求提交(推荐): 飞书表单 https://mcnnox2fhjfq.feishu.cn/share/base/form/shrcnkR1s45VDuFnV3GbD6VhnIJ
GitHub: Issues https://github.com/CherryHQ/cherry-studio/issues | Discussions https://github.com/CherryHQ/cherry-studio/discussions | 看板 https://github.com/orgs/CherryHQ/projects/7
社群: Discord https://discord.gg/wez8HtpxqQ | Telegram https://t.me/CherryStudioAI | X https://twitter.com/CherryStudioHQ | QQ群 575014769 | 论坛 linux.do
官网: cherry-ai.com | 中文文档 docs.cherry-ai.com | 主题 cherrycss.com | 邮箱 [email protected] / [email protected]
中文用户推荐QQ群/linux.do/飞书表单, 国际用户推荐Discord/Telegram/GitHub
提交Issue前检测 gh auth status。未登录→告知安装 https://cli.github.com/ 后 gh auth login。不想配→记录本地+引导飞书表单/社区论坛
macOS正式: ~/Library/Application Support/CherryStudio/logs/ | 开发: CherryStudioDev/logs/ | Windows: %APPDATA%/CherryStudio/logs/
Production-ready reinforcement learning algorithms (PPO, SAC, DQN, TD3, DDPG, A2C) with scikit-learn-like API. Use for standard RL experiments, quick prototyping, and well-documented algorithm implementations. Best for single-agent RL with Gymnasium environments. For high-performance parallel training, multi-agent systems, or custom vectorized environments, use pufferlib instead.
Create publication-quality scientific diagrams using Nano Banana 2 AI with smart iterative refinement. Uses Gemini 3.6 Flash for quality review. Only regenerates if quality is below threshold for your document type. Specialized in neural network architectures, system diagrams, flowcharts, biological pathways, and complex scientific visualizations.
Implement semantic vector search with AgentDB for intelligent document retrieval, similarity matching, and context-aware querying. Use when building RAG systems, semantic search engines, or intelligent knowledge bases.
Context is the complete state available to a language model at inference time. It includes everything the model can attend to when generating responses: system instructions, tool definitions, retrieved documents, message history, and tool outputs.
Build semantic vector search systems with AgentDB for intelligent document retrieval, RAG applications, and knowledge bases using embedding-based similarity matching
Implement semantic vector search with AgentDB for intelligent document retrieval, similarity matching, and context-aware querying. Use when building RAG systems, semantic search engines, or intelligent knowledge bases.
Add a persistent wiki knowledge base to a NanoClaw group. Based on Karpathy's LLM Wiki pattern. Triggers on "add wiki", "wiki", "knowledge base", "llm wiki", "karpathy wiki".
Search the ExecuTorch tribal knowledge base covering QNN, XNNPACK, Vulkan, CoreML, Arm, and Cadence backends, quantization recipes, export pitfalls, runtime errors, and SoC compatibility. Use when debugging ExecuTorch errors, choosing quantization configs, checking backend op support, or answering questions about Qualcomm HTP / Snapdragon / Apple Neural Engine behavior.
Take cherryhq/cherry-assistant-guide from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.