Skills
skills/agrimsingh/ralph-wiggum-cursor/ralph-wiggum/Gen Agent Trust Hub

ralph-wiggum

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • EXTERNAL_DOWNLOADS (HIGH): The installer script install.sh downloads multiple bash scripts from an untrusted personal GitHub repository (agrimsingh/ralph-wiggum-cursor) to the local .cursor/ralph-scripts/ directory.
  • REMOTE_CODE_EXECUTION (HIGH): Multiple files, including install.sh and scripts/init-ralph.sh, contain instructions to download and execute scripts directly from the internet using the curl | bash pattern. This includes downloads from cursor.com and raw.githubusercontent.com, neither of which are within the specified trusted organization scope for this audit.
  • COMMAND_EXECUTION (HIGH): The install.sh script employs sudo to modify system-level package repositories and install the gum utility on Debian, Fedora, and RHEL-based systems, which is an unnecessary elevation of privilege for a developer tool skill.
  • INDIRECT_PROMPT_INJECTION (LOW): The skill has a large attack surface as it ingests untrusted task definitions from RALPH_TASK.md (Ingestion point) without boundary markers or sanitization. The agent possesses capabilities to execute arbitrary shell commands via the test_command property and perform git operations (Capability inventory), creating a vector where malicious instructions in a task file could lead to unauthorized code execution.
Recommendations
  • HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/agrimsingh/ralph-wiggum-cursor/main/install.sh, https://cursor.com/install - DO NOT USE without thorough review
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 05:47 PM