Back to plugin

Security audit

Feishu/Lark

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent Feishu/Lark integration, but it exposes powerful cloud document, drive, wiki, and table actions while some skill triggers are too broad.

Review before installing if this Feishu/Lark app will have broad workspace scopes. Configure least-privilege Feishu app permissions, keep the permission-management tool disabled unless needed, use explicit Feishu/Lark links or IDs, and require confirmation for deletes, moves, permission changes, whole-document writes, and uploads of private local files or remote media.

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The documentation states that only list_blocks, update_block, and delete_block should be used, but later instructs use of additional actions such as upload_image, create_table, insert, and other table operations. This inconsistency can mislead an agent into invoking undocumented or unsupported tool capabilities, increasing the chance of unsafe writes, unexpected behavior, or bypass of intended tool constraints.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The activation description is broad enough to trigger on ordinary mentions of a 'knowledge base', 'wiki', or 'wiki link', which can cause the skill to activate outside the user's actual intent. Over-broad activation increases the chance of unintended tool use, unnecessary exposure of accessible Feishu wiki metadata, or confusion from the agent taking actions in the wrong context.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The activation description includes broad, common terms like 'folders', 'Drive files', and 'document comments', which can match many ordinary user requests and cause the skill to activate when the user did not specifically intend Feishu Drive access. In a skill that can access shared cloud-storage contents and comments, over-broad invocation increases the chance of unnecessary data exposure or unintended operations.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The activation text says to activate when the user mentions "Feishu docs, cloud docs, or docx links." While "Feishu docs" is specific, "cloud docs" and especially "docx links" are broad phrases that could match many unrelated conversations and do not clearly define scope or exclusions. The file also provides no negative examples or constraints to limit unintended invocation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

This markdown file instructs users to use delete_block to remove document blocks, which is a destructive operation affecting user data. While it mentions confirming broad deletions, it does not clearly warn that deleting the wrong block IDs may permanently remove content or advise caution for any deletion operation.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.