ClawHub
Back to skill

Security audit

Iflytek Hyper Tts

Security checks across malware telemetry and agentic risk

Overview

This is a straightforward cloud text-to-speech skill, but users should understand that text is sent to iFlytek and audio files are saved locally.

Install only if you intend to use iFlytek's cloud TTS service. Do not synthesize secrets, private customer data, regulated content, or confidential drafts unless sending that text to iFlytek is approved. Keep XFEI credentials in environment variables or a secret manager, and choose output paths deliberately because generated audio is saved to disk.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Intent-Code Divergence

Medium
Confidence
84% confidence
Finding
The documentation says only 7 curated voices are supported, but nearby examples and parameter descriptions allow arbitrary VCN values, including voices outside that set. This inconsistency can bypass intended restrictions, causing unreviewed or unauthorized voice selection and weakening any safety or product constraints tied to the curated list.

Intent-Code Divergence

Medium
Confidence
82% confidence
Finding
The main documentation presents MP3/lame as the supported output format, while the FAQ advertises several additional encodings. Conflicting interface claims can cause downstream systems or agents to make unsafe assumptions about file type handling, validation, storage, or processing paths.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The README instructs users to send arbitrary text to a third-party cloud TTS API and configure API credentials, but it does not warn that submitted text may leave the local environment and be processed by an external service. This can lead to inadvertent disclosure of sensitive or regulated content if users assume synthesis happens locally or do not understand the data-sharing implications.

Vague Triggers

Medium
Confidence
79% confidence
Finding
The trigger conditions include broad, everyday phrases such as requests to 'read this text aloud,' which may match normal conversation too loosely. Overbroad activation can unintentionally send user-provided content to an external service and generate files without sufficiently clear user intent.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill description explains text-to-speech functionality but does not clearly warn that supplied text is transmitted to an external provider and saved as an output file. This is a meaningful privacy and data-handling omission, especially when users may provide sensitive text expecting only local transformation.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
scripts/xfei_hyper_tts.py:198