The Agent Skills Directory

[SAFE]: The skill employs a local Python script (scripts/check_copy.py) to enforce specific writing style and formatting rules (e.g., banning certain phrases, checking word counts). This script relies exclusively on the Python standard library and does not perform any network operations, external downloads, or hazardous system calls.
[SAFE]: The agent's instructions include explicit privacy constraints, forbidding the disclosure of private figures, client identities, or sensitive business metrics. These constraints are reinforced by specific evaluation cases in evals/cases.json designed to verify that the agent deflects personal inquiries.
[DYNAMIC_EXECUTION]: The test file tests/test_check_copy.py uses importlib to dynamically load and execute the skill's own local linter script for unit testing. As this execution is scoped to the skill's internal files and occurs within a testing context, it does not pose a security risk.
[SAFE]: No evidence of prompt injection, data exfiltration, obfuscation, or persistence mechanisms was found during analysis.

fede-clone