The Agent Skills Directory

[SAFE]: The logic implemented in scripts/verify_edit.py is entirely deterministic and relies on Python standard library modules for text processing and statistics. It avoids high-risk operations like network communication, subprocess creation, or dynamic code evaluation.\n- [SAFE]: The skill processes local text files provided by the user to perform multiset equality checks. It does not possess capabilities to write to the file system or transmit data to external servers, and it does not interpret processed text as instructions for the agent.\n- [SAFE]: The documentation and external references point to legitimate academic sources for documenting the verification metrics. No signs of prompt injection, obfuscation, or persistence mechanisms were detected.

deslop-verify