Back to skill

Security audit

feishu-message-group-whitelist

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple Feishu group-message keyword whitelist with no executable code, network access, hidden persistence, or credential handling.

Install only if you want Feishu group bot replies gated by local keyword matching. Review config/whitelist.txt before use, and avoid leaving the whitelist empty unless you intentionally want the bot to reply without filtering.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest description lists trigger examples such as '群里有人发消息时' and '需要过滤无关消息时', which are broad everyday conditions rather than specific invocation constraints. This makes it unclear when the skill should be considered applicable versus not applicable, increasing the risk of unintended activation.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

If the whitelist is empty, the documented behavior disables filtering and allows the bot to respond to all group messages. In a group-chat skill whose purpose is to restrict activation, this fail-open design can unexpectedly turn a narrowly scoped bot into a broadcast-responsive one, increasing unintended data exposure, noisy automation, and abuse surface.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The whitelist contains only specific language/script entries, including a Chinese term and a Latin-script token, which suggests a locale or language restriction may be enforced. Because this file provides no natural-language justification or indication of user choice, it may violate the policy against forcing a specific language or locale without opt-in.

Content

No source excerpt is available for this finding.