skills/rheinmir/setup/harness-tour/Gen Agent Trust Hub

harness-tour

Fail

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill provides a command for the user to install a dependency using curl -fsSL https://raw.githubusercontent.com/Rheinmir/setup/orca/harness/poc-vendor-neutral/bootstrap.sh | bash. Executing remote scripts by piping them directly to a shell is a dangerous pattern that allows for arbitrary code execution on the user's system.
  • [COMMAND_EXECUTION]: The skill performs multiple system checks and executions, including environment detection using grep and test, and executing local scripts such as llmwiki-validate.py, harness-events.py, and demo.sh. It also instructs the agent to perform file writes and deletions to simulate rule violations.
  • [EXTERNAL_DOWNLOADS]: The skill attempts to fetch configuration and installation scripts from an external source (raw.githubusercontent.com). While the source matches the skill author's repository, the method of fetching and immediate execution remains a significant security concern.
Recommendations
  • HIGH: Downloads and executes remote code from: https://raw.githubusercontent.com/Rheinmir/setup/orca/harness/poc-vendor-neutral/bootstrap.sh - DO NOT USE without thorough review
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jul 15, 2026, 08:40 AM
Security Audit — agent-trust-hub — harness-tour