T09 · Insecure Skill Coding Practices
Error
- Location
scripts/voice-bubble.mjs:65- Finding
Arbitrary Command Execution Through Shell Command Injection
- Content
View full analysis
- Remediation
View remediation
Security audit
Feishu Voice Bot
Security checks for vulnerabilities and agentic risk
Overview
This Feishu voice-message skill has a coherent purpose, but its helper script can turn message text or options into shell command execution.
Review before installing. The Feishu voice-bubble behavior is disclosed, but the helper script should be patched to use execFileSync or spawnSync with argument arrays, validate voice/rate/output values, avoid silent overwrites, and clean temporary files before use in any environment where message text or options may be influenced by others.
Vulnerability Patterns
- Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
- 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
Findings (1)
Vulnerability Patterns
- Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
- MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
- 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 (6)
Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md node scripts/voice-bubble.mjs "你好陛下" --voice zh-CN-XiaoxiaoNeural
Undeclared Tool Scope
Medium
- Category
- MCP Least Privilege
- Confidence
- 70% confidence
- Finding
Without declared permissions the skill's intent is opaque and cannot be validated.
- Content
Vague Triggers
Medium
- Category
- Not specified by scanner
- Confidence
- 92% confidence
- Finding
The activation guidance includes broad conditions such as any time the user wants TTS or voice output, which can cause the skill to be selected outside the intended Feishu-specific context. That can lead to unintended message sending, audio generation, or use of external tooling when a simpler or safer response mode was expected.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 84% confidence
- Finding
The script invokes ffmpeg with the
-yflag, which force-overwrites the output file if it already exists. Although the file's general conversion behavior is documented, there is no explicit warning or confirmation that a user-specified--outputpath may be overwritten irreversibly.- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 90% confidence
- Finding
The documentation states that the default voice is
zh-CN-XiaoxiaoNeural, which imposes a specific language/locale choice by default. The file does list supported English voices, but it does not say that the user will be asked to choose or that the Chinese default is only used in a documented region-specific context.- Content
Natural-Language Policy Violations
Low
- Category
- Not specified by scanner
- Confidence
- 75% confidence
- Finding
The usage text and defaults are centered on Chinese text and a Chinese voice locale (
zh-CN-XiaoxiaoNeural), which can imply a fixed language/locale choice. The file does allow--voiceoverrides, but it does not explicitly present language choice as an opt-in or clarify that other locales are equally supported.- Content