Skip to content

feat: v2 — 飞书全能集成套件 (CLI + MCP Integration Kit)#4

Open
AlexAnys wants to merge 1 commit into
mainfrom
v2-integration-kit
Open

feat: v2 — 飞书全能集成套件 (CLI + MCP Integration Kit)#4
AlexAnys wants to merge 1 commit into
mainfrom
v2-integration-kit

Conversation

@AlexAnys
Copy link
Copy Markdown
Owner

@AlexAnys AlexAnys commented Mar 31, 2026

Summary

将 feishu-mcp 从 MCP-only 工具转型为飞书全能集成套件 — 让飞书 CLI(200+ 命令、19 个 AI Skills)和社区 MCP(15 个文档编辑工具)作为统一系统工作的集成层。

核心定位

  • 不是 MCP 服务器,不是 CLI 替代品 — 是让两个工具协同工作的集成套件
  • CLI 负责广度(11 个业务域:消息、日历、邮件、任务、通讯录、会议、Wiki、云盘、表格、多维表格)
  • 社区 MCP 负责深度(文档 Block 级编辑、批量创建、精细格式控制)

新增内容(+4726 行,16 个文件)

安装自动化

  • setup-v2.sh — 智能安装脚本:一条命令安装 CLI + MCP + OpenClaw 检测
  • configs/ — 3 个平台配置模板(OpenClaw / Claude Code / Cursor)

AI Agent 文档

  • SKILL.md — 544 行决策矩阵(49 个意图 → 工具路由)
  • AGENTS.md — 749 行 Agent 操作手册 + 15 入口决策树 + 39 个命令示例
  • QUICKSTART-AGENT.md — 9 步自动安装流程
  • TROUBLESHOOT-AGENT.md — 17 个错误条目 + 恢复步骤

用户文档

  • README.md — 重写为非技术用户视角
  • GUIDE.md — 从零开始 7 步配置指南 + 10 个示例对话

工作流 Skill

  • skills/feishu-overview.md — 能力总览索引
  • skills/doc-review.md — 文档审阅(MCP 读 → CLI 建任务 → CLI 发群)
  • skills/data-to-doc.md — 数据报告(CLI 查表 → MCP 生成格式化文档)
  • skills/knowledge-sync.md — 知识同步(Wiki ↔ 文档双向)
  • skills/meeting-summary-plus.md — 增强会议纪要

开发方法

使用 Planner → Builder → QA 三角色 harness:

  • Phase 1(安装脚本):QA 发现 2 个 blocker → 修复后 PASS
  • Phase 2(决策矩阵+Agent 文档):QA 评分 8.35/10 PASS
  • Phase 3-4(用户文档+工作流 Skill):集成检查 PASS

Test plan

  • 在干净环境运行 setup-v2.sh,验证 CLI 安装和 auth
  • 验证 SKILL.md 决策矩阵:50 个常见飞书操作意图的路由准确率
  • 非技术用户按 GUIDE.md 操作能否 10 分钟内完成配置
  • 每个 workflow skill 的 CLI + MCP 组合是否正常工作
  • 所有 skill 文件能被 npx skills add 正确识别

Transform feishu-mcp from MCP-only tool into a comprehensive integration
kit combining Feishu CLI (200+ commands, 19 skills) with community MCP
(15 doc editing tools) for non-technical users.

New deliverables:
- setup-v2.sh: smart setup script with CLI + MCP install, OpenClaw detection
- SKILL.md: 544-line decision matrix (49 intents → tool routing)
- AGENTS.md: 749-line master agent reference with decision tree
- QUICKSTART-AGENT.md: 9-step automated setup guide for AI agents
- TROUBLESHOOT-AGENT.md: 17 error entries with recovery procedures
- GUIDE.md: step-by-step visual setup guide for non-technical users
- 3 platform config templates (OpenClaw, Claude Code, Cursor)
- 5 workflow skills demonstrating CLI + MCP synergy

Architecture: two-layer (CLI for breadth, community MCP for depth).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Copy link
Copy Markdown

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 0e21fe5a40

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

  • Open a pull request for review
  • Mark a draft as ready
  • Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

Comment thread setup-v2.sh
# ============================================================

SETUP_VERSION="2.0.0"
SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Handle stdin execution when resolving SCRIPT_DIR

The setup flow is documented to run as curl .../setup-v2.sh | bash, but in that execution mode $0 is bash, so SCRIPT_DIR resolves to the caller’s current directory rather than the repository/script location. Later steps read templates from ${SCRIPT_DIR}/configs/*.json to generate Claude/Cursor/OpenClaw configs, which will be missing for piped installs, so users following the documented install path won’t get platform config generation.

Useful? React with 👍 / 👎.

Comment thread setup-v2.sh
Comment on lines +376 to +378
if stty -echo 2>/dev/null; then
read -r APP_SECRET
stty echo
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Ensure terminal echo is restored on secret prompt abort

The secret prompt disables terminal echo with stty -echo, but if read is interrupted (for example, user presses Ctrl+C) the script exits before stty echo runs, leaving the user’s terminal in no-echo mode for subsequent commands. This is an interactive reliability issue and is especially visible during manual setup when users cancel at the secret entry step.

Useful? React with 👍 / 👎.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant