detecting-port-conflicts
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use standard system utilities
lsofandkillto identify and terminate processes occupying specific network ports. These commands are used for their intended purpose in a developer troubleshooting context. - [INDIRECT_PROMPT_INJECTION]: The skill operates by scanning terminal output for error strings like
EADDRINUSE. This creates an ingestion surface where external application output could trigger the agent's diagnostic logic. However, the resulting actions are limited to port discovery and process management, and the skill explicitly requires user confirmation before performing broad actions likekillall node. - [PRIVILEGE_ESCALATION]: The skill notes that ports below 1024 require root access on macOS, but it does not provide commands to automatically escalate privileges (e.g., using
sudo). It leaves privilege management to the user's environment.
Audit Metadata