mathodology-whole-project

Fail

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The SKILL.md file contains instructions for the agent to download a Python script from a remote GitHub repository and execute it locally (curl | python3). This bypasses standard package management verification.
  • [EXTERNAL_DOWNLOADS]: The scripts/update-project.py script performs network operations to fetch source archives from codeload.github.com and resolves repository references via the GitHub API. It also utilizes npx and uvx to download and run software packages during execution.
  • [DYNAMIC_EXECUTION]: The scripts/update-project.py script includes a hardcoded Python supervisor script that is executed at runtime via sys.executable -c to manage process lifecycles and timeouts.
  • [COMMAND_EXECUTION]: The skill executes multiple shell commands for file management, process termination, and environment configuration, including the use of git, tar, and chmod in its backup and update scripts.
Recommendations
  • HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/sweetcornna/mathodology/main/.claude/skills/mathodology-whole-project/scripts/update-project.py - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Sep 7, 2026, 02:40 PM
Security Audit — agent-trust-hub — mathodology-whole-project