pencil-mcp-batch-get

name	pencil-mcp-batch-get
description	Batch search and read node information. The Agent's 'Eyes'. Use to find specific components e.g. all nodes named 'Button' or get child structure within a container.
license	Complete terms in LICENSE.txt

Tools

This skill is designed to call the Pencil MCP tool:

• batch_get

If your client namespaces MCP tools, it may appear as mcp__pencil__batch_get.

When to use this skill

Intent Recognition (CRITICAL)

Even if a trigger phrase matches, you must verify the user's intent:

1. Is the user explicitly asking to use "Pencil"?
2. Is the current conversation context clearly about "Pencil" design tasks?

If the answer is NO, do NOT use this skill. (e.g., if the user just says "Find the button" in a general context, they might mean simple text search or other tools).

CRITICAL PREREQUISITE: You must ONLY use this skill when the user EXPLICITLY mentions "Pencil".

ALWAYS use this skill when:

• You need to Find specific nodes by name or ID.
• You need to Read the properties of specific nodes.
• You need to inspect the children of a Frame.
• The user asks to "Find the button", "Get properties of X" in Pencil.

Trigger phrases include:

• "Pencil find all buttons" (Pencil 查找所有按钮)
• "Get node info with Pencil" (用 Pencil 获取节点信息)
• "Read properties using Pencil" (使用 Pencil 读取属性)
• "Pencil search components" (Pencil 搜索组件)

Input Parameters

• filePath (string, optional): Path to file.
• patterns (array, optional): Search patterns (e.g., ["name=Button"]).
• nodeIds (array, optional): Specific IDs to read.
• searchDepth (integer, optional): Depth for search.
• readDepth (integer, optional): Depth for reading children structure. Keep low (<3).
• includePathGeometry (boolean, optional): Include vector path data.
• resolveVariables (boolean, optional): Return computed values instead of variable refs.

How to use this skill

1. Combine Requests: If you need to search AND read by ID, do it in ONE call.
2. Smart Traversal:
   • Start with top-level or known IDs.
   • If you see ... (truncated children), make a new call with those specific child IDs.
3. Design Systems: To list available components, search for reusable nodes inside the design system frame.

Examples

1. Simple: Get Root Children

Get the top-level nodes of the document to understand the general structure. See 1-get-root.json.

2. Medium: Search by Name

Find all nodes that contain "Button" in their name. See 2-search-name.json.

3. Complex: Detailed Search

Search for specific nodes by ID and name pattern, resolving variables to see actual values, and getting full geometry. See 3-detailed-search.json.

Keywords

English keywords: batch get, search nodes, find elements, read properties, inspect structure, get components

Chinese keywords (中文关键词): 批量获取, 搜索节点, 查找元素, 读取属性, 检查结构, 获取组件

能力边界

✅ 适用场景

• 当你需要使用此技能对应的技术栈时
• 当项目需要遵循最佳实践时
• 当需要快速上手或深入理解核心概念时

⚠️ 需要注意

• 复杂业务逻辑需要结合具体场景调整
• 性能优化需要根据实际数据量评估

❌ 不适用场景

• 不相关的技术栈或框架
• 需要完全自定义的特殊场景

常见陷阱 (Gotchas)

1. 版本兼容性：注意框架版本与依赖库的兼容性，不同版本 API 可能有差异
2. 配置文件格式：配置文件格式错误是最常见的问题，建议使用编辑器的语法检查
3. 环境变量：确保所有必要的环境变量已正确设置，敏感信息不要硬编码
4. 依赖冲突：多版本共存时注意依赖冲突，使用 lock 文件锁定版本
5. 性能陷阱：大数据量场景下注意性能优化，避免 N+1 查询等常见问题

使用流程

Step 1: 环境准备

确保开发环境已安装必要的依赖和工具。

Step 2: 配置初始化

根据项目需求进行基础配置。

Step 3: 核心功能使用

按照示例代码实现核心功能。

Step 4: 测试验证

运行测试确保功能正常。

Step 5: 部署上线

完成开发后进行部署和监控。