mcpField
F
fde-mock-knowledge-base

fde-mock-knowledge-base

github.com/letianpaicyDocuments143 installsUpdated 14w ago
Repository

About fde-mock-knowledge-base

A skill that creates six sanitized mock beauty-retail knowledge documents in Feishu: five source documents and one index. It is intended for training exercises and provides source material without test questions, expected answers, or evaluation results.

Key features

  1. 01Create five mock source documents
  2. 02Create one index document
  3. 03Preview creation with a dry run
  4. 04Require confirmation before writing
  5. 05Return created document links

Use cases

  1. 01Prepare mock retail material for a training exercise
  2. 02Create a sample document set in Feishu
  3. 03Provide source content without test answers

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/letianpaicy/watcha-fde --skill fde-mock-knowledge-base

After installing, type/fde-mock-knowledge-basein Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

agents/
openai.yaml
assets/beauty-retail/
01-product-label.md
02-inventory-expiry.md
03-claims-consulting.md
04-return-complaint.md
05-adverse-reaction.md
references/
data-schema.md
lark-setup.md
scripts/
create_feishu_docs.py
preflight.py

Details

From the project README

FDE 美妆零售知识库 Mock

创建 5 篇脱敏模拟美妆零售知识文档和 1 篇目录索引。只提供原始业务材料,不生成测试问题、预期答案或验收结果。

工作流

  1. 将包含本文件的目录记为 SKILL_DIR。
  2. 运行 python3 "$SKILL_DIR/scripts/preflight.py"。
  3. 按 飞书环境处理 处理预检结果;环境未就绪时不要继续。
  4. 不询问行业、角色、文档目录、规模或测试集,直接运行预览:
python3 "$SKILL_DIR/scripts/create_feishu_docs.py" --dry-run
  1. 告诉用户将直接在“我的文档库”创建 6 篇美妆零售新文档,不修改或删除现有数据,并请求一次写入确认。
  2. 用户确认后运行:
python3 "$SKILL_DIR/scripts/create_feishu_docs.py" --yes
  1. 返回脚本输出中的目录和文档链接;明确材料是脱敏模拟内容,不代表真实企业制度或专业意见。

约束

  • 不要替学员生成知识库测试问题、预期答案、是否通过或准确率。
  • 不要要求用户提供目标文件夹、Wiki Token 或文档目录。
  • 不要提供制造业或医药材料,也不要根据用户身份改写预置材料;目标角色由学员在 Day 1 自行定义。
  • 写入部分失败时,报告已创建链接和错误;不要自动删除任何飞书资源。
  • 只使用脚本提供的公开来源整理材料,不抓取或复制报告全文。

资源

Comments

Markdown supported