product-changelog

Fail

Audited by Gen Agent Trust Hub on Feb 18, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION] (CRITICAL): Detected piped remote execution: curl -fsSL https://cli.inference.sh | sh. This command downloads a script from an unverified remote source and executes it immediately in the shell, which is a major security risk for remote code execution (RCE).\n- [EXTERNAL_DOWNLOADS] (HIGH): The skill references and executes code from untrusted sources via inference.sh and npx skills. These domains and packages are not included in the list of Trusted External Sources, posing a supply-chain risk.\n- [COMMAND_EXECUTION] (MEDIUM): The skill configuration allowed-tools: Bash(infsh *) grants the agent permission to execute any shell command starting with infsh. This provides a broad capability that could be abused to interact with the local system.\n- [PROMPT_INJECTION] (LOW): The skill processes untrusted product descriptions and changelog data (Ingestion: SKILL.md) to generate language. It lacks explicit boundary markers or sanitization, creating a surface for indirect prompt injection that could influence agent behavior.
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
Feb 18, 2026, 01:15 AM