飞书多维表格 AI 管家

• Outcome: Review
• Latest audit: May 29, 2026 at 9:39 PM UTC
• Version:
  1.0.2

Overview

This Feishu Bitable helper is mostly coherent, but it gives agents broad read/write workflows for business, customer, and employee tables without enough built-in confirmation or scoping guidance.

Install only if you intend to let an agent operate on Feishu Bitable data. Limit the Feishu app to the specific bases and tables needed, require a preview before any bulk read/write or field change, avoid sending summaries to chat unless explicitly approved, and treat customer or employee tables as sensitive production data.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Context-Inappropriate Capability

Medium

• Confidence: 91% confidence
• Finding: The example explicitly adds `feishu_chat` group messaging as an optional step, which extends the skill from Bitable data management into outbound communications. That scope expansion can enable unintended data disclosure, especially if summarized table contents or sensitive business metrics are posted to a chat without clear authorization or user confirmation.

Vague Triggers

Medium

• Confidence: 89% confidence
• Finding: The trigger list includes broad, commonplace terms such as “数据清洗” and “飞书数据”, which can cause the skill to activate in conversations that are not specifically requesting Bitable operations. Because this skill can read table contents and perform writes, accidental invocation increases the chance of unintended access to sensitive data or unintended record modifications.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The skill documentation encourages reading records, traversing entire tables, generating summaries, and writing results back, but it does not instruct the agent to warn the user about privacy, scope, or destructive-change risks beforehand. In a Bitable context, these actions may expose personal or business-sensitive information and may alter production data, so missing consent and risk disclosure materially increases the chance of harmful misuse or accidental overreach.

Missing User Warnings

Medium

• Confidence: 85% confidence
• Finding: The examples show bulk employee record updates and additions without any warning, confirmation, or authorization guidance for modifying personnel data. In practice, this can normalize high-impact administrative actions and increase the chance of accidental or unauthorized changes to sensitive HR records.

VirusTotal

65/65 vendors flagged this skill as clean.