Loading ToolBench
飞书 CLI → MCP Bridge：将 lark-cli 能力以 MCP tools 形式暴露给 AI Agent
Server has 11 tools with basic descriptions and parameter schemas. However, critical gaps undermine quality: (1) Parameter descriptions are minimal or missing context (e.g., 'text' param in im_messages_send lacks guidance on length/format); (2) No output schemas documented, LLMs cannot predict response structure; (3) Error handling is generic JSON with error_code/message but lacks recovery guidance or actionable next steps; (4) Descriptions are functional but lack LLM-optimized context (e.g., 'lark.identity' is 16 chars, below the 50-200 char baseline); (5) No tool annotations (readOnlyHint, destructiveHint) despite clear risk levels (WRITE vs READ_ONLY); (6) Mutually exclusive parameters (chat_id vs user_id, url vs spreadsheet_token) documented in descriptions but not enforced via schema constraints. Baseline: 549 production tools average 194 chars for descriptions and 72 chars for params, this server averages ~80 chars for descriptions and ~60 for params.
搜索多维表格记录。根据关键词在指定表格中搜索。[user/bot 身份均可]
查看日历日程，默认展示今天的安排。可指定日期范围。[需要 user 身份]
搜索飞书联系人。支持按姓名、邮箱等关键词搜索用户。[需要 user 身份]
读取飞书文档内容。支持通过文档 URL 或 token 获取。[需要 user 身份]
查看所有可用的飞书域及各域 tool 数量概览。
查看当前飞书登录身份（用户/机器人）、状态、open_id 等信息。
发送消息到指定会话或用户。支持 text、markdown、post、image 等消息类型。chat_id 和 user_id 二选一。[user/bot 身份均可]
No output schemas documented. LLMs cannot predict response structure or plan downstream tool calls. All 11 tools return JSON but structure is undocumented.
Tool descriptions are too short (avg 80 chars vs baseline 194). Descriptions like 'lark.identity' (16 chars) lack context for LLM selection. No guidance on when to use vs similar tools.
Missing tool annotations despite clear risk levels. Tools marked WRITE (im_messages_send, sheets.write) lack destructiveHint=true. READ_ONLY tools lack readOnlyHint=true. Agents cannot infer safety properties.
Inferred effective spec: 2025-06-18+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 53 | 2025-06-18+ | v2 |
查看当前飞书应用已授权的 scope 列表。用于判断哪些操作可执行。
读取电子表格数据。支持通过 URL 或 token + range 指定读取范围。[需要 user 身份]
写入电子表格数据（覆盖模式）。支持通过 URL 或 token + range 指定写入范围。[需要 user 身份]
查看分配给我的任务列表。支持按完成状态、截止日期、关键词过滤。[需要 user 身份]
Error handling is generic. _format_result() returns {error_code, message, recovery_hint} but recovery_hint is hardcoded ('此命令被安全策略禁止。详见 SECURITY.md'). No actionable guidance for execution errors (API failures, validation errors).
Mutually exclusive parameters (chat_id vs user_id, url vs spreadsheet_token) documented in text but not enforced via schema. LLMs may pass both, causing ambiguous failures.