atlassian-with-mcporter
Warn
Audited by Gen Agent Trust Hub on May 11, 2026
Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill hardcodes absolute file paths to a local binary (
/Users/tao.exe/.nvm/versions/node/v22.22.2/bin/mcporter) within its instructions. This exposes local environment metadata and relies on specific system configurations that may not be present or secure across different environments. - [COMMAND_EXECUTION]: The skill includes instructions to bypass security boundaries. Specifically, it advises the agent to suggest rerunning the
mcportercommand outside of the restricted sandbox environment if a permission error (EPERM) is encountered, which directly encourages the circumvention of execution isolation policies.
Audit Metadata