The Agent Skills Directory

[SAFE]: The skill operates as a local log analysis utility as described. It uses standard Python libraries (argparse, json, re, datetime) to parse log files from the user's home directory (~/.openclaw/logs/) and temporary storage (/tmp/openclaw/). The analysis found no network operations, hardcoded credentials, or persistence mechanisms. While the script processes log data which could technically contain arbitrary strings (Indirect Prompt Injection surface), it lacks any dangerous capabilities like code execution or network exfiltration, and the operation is restricted to the specific log formats and paths associated with the OpenClaw application.

openclaw-logs