The Agent Skills Directory

[SAFE]: The skill uses established read-only tools (harness_list, harness_describe) to query audit events. There is no evidence of command execution, network exfiltration, or persistence mechanisms.
[SAFE]: No external packages or remote scripts are downloaded; all report templates are provided as static markdown within the skill's reference files.
[SAFE]: The instructions focus on summarizing existing log data into structured reports without attempting to bypass security constraints or override agent behaviors.

audit-report