The Agent Skills Directory

[SAFE]: The skill is composed entirely of instructional markdown and metadata. It contains no executable scripts, shell commands, or external package dependencies.
[PROMPT_INJECTION]: No malicious injection patterns or safety bypass attempts were detected. The instructions prioritize accuracy by mandating verification of legal status via specific connectors before providing advice.
[DATA_EXFILTRATION]: The skill does not perform network operations or access sensitive file paths. The allowed-tools configuration is restricted to Read, and pii-egress is set to none.
[COMMAND_EXECUTION]: There are no patterns involving sudo, chmod, or shell execution. It relies on a 'human-in-the-loop' model where the final output is a draft for human review rather than an automated action.
[SAFE]: The skill follows security best practices for AI agents, including requiring human approval for outputs and providing clear disclaimers that it does not replace technical audits or legal counsel.

nis2-ksc-pl