ANY WHISPER API

Security checks across malware telemetry and agentic risk

Overview

This transcription skill is purpose-aligned, but it should be reviewed because it uploads audio and an API key to a configurable server while handling that server address unsafely.

Review before installing. Use only a WHISPER_API_HOST you trust, preferably a local or known server, because it receives the audio content and bearer token. Avoid sensitive audio until the host is quoted and validated and the skill adds clearer upload/privacy warnings.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
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)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 70% confidence
Finding: Without declared permissions the skill's intent is opaque and cannot be validated.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: This markdown file describes transcribing audio via an API endpoint and notes that output defaults to a local text file, but it does not include any user-facing warning about privacy or data-handling implications. Because audio content may contain sensitive information and the skill transmits it over HTTP and stores derived text on disk, the description should disclose those behaviors.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: This shell script uploads the provided audio file to an external HTTP endpoint using curl, which is a privacy-relevant network operation. While the script name and arguments imply transcription, there is no explicit disclosure in comments, usage text, or runtime output that the audio will be sent to a remote service.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal