skills/narasaka/skills/dotenvx/Gen Agent Trust Hub

dotenvx

Fail

Audited by Gen Agent Trust Hub on Nov 12, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill documentation includes installation examples that pipe remote shell scripts to the system shell. Evidence: curl -sfS https://dotenvx.sh | sh and curl -fsS https://dotenvx.sh/install.sh | sh found in the Installation, Docker, and CI/CD sections.
  • [INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface through project environment files. Ingestion points: .env and .env.keys files. Boundary markers: The -- separator is documented for command execution but does not provide validation for the content being injected. Capability inventory: Primary capability is the execution of arbitrary system processes via dotenvx run. Sanitization: No sanitization logic for environment file content is described.
  • [COMMAND_EXECUTION]: The tool's intended use involves executing arbitrary commands and scripts, such as dotenvx run -- node index.js.
  • [DYNAMIC_EXECUTION]: The skill describes support for shell command substitution ($(command)) within environment files, allowing for the execution of arbitrary commands defined in configuration data at runtime.
Recommendations
  • HIGH: Downloads and executes remote code from: https://dotenvx.sh, https://dotenvx.sh/install.sh - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Nov 12, 2026, 04:03 AM