Skills
skills/tul-sh/skills/ai-voice-cloning/Gen Agent Trust Hub

ai-voice-cloning

Fail

Audited by Gen Agent Trust Hub on Mar 4, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill recommends installing the infsh CLI tool using curl -fsSL https://cli.inference.sh | sh. This pattern is highly insecure as it executes a remote script from a third-party domain without local verification or auditing.\n- [EXTERNAL_DOWNLOADS]: The skill documentation mentions downloading binaries and checksums from dist.inference.sh. It also suggests adding related skills via npx skills add inference-sh/skills@..., which fetches external code at runtime.\n- [COMMAND_EXECUTION]: The skill's primary functionality relies on executing the infsh binary through the Bash tool. The security of these commands depends entirely on the integrity of the downloaded binary and the remote environment it interacts with.\n- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by interpolating user-provided text into shell commands for the infsh tool. There are no boundary markers or sanitization logic present to prevent embedded instructions in the text from influencing the tool's behavior.\n
  • Ingestion points: The --input JSON payload in multiple bash examples (e.g., SKILL.md).\n
  • Boundary markers: Absent.\n
  • Capability inventory: Access to the Bash tool for running infsh commands.\n
  • Sanitization: Absent.
Recommendations
  • HIGH: Downloads and executes remote code from: https://cli.inference.sh - DO NOT USE without thorough review
  • AI detected serious security threats
Audit Metadata
Risk Level
CRITICAL
Analyzed
Mar 4, 2026, 12:29 PM