The Agent Skills Directory

[SAFE]: The skill defines a specific historical persona using detailed personality guidelines and concept definitions provided in the included reference.md file.
[SAFE]: Persona initialization is managed through local references and a repository-local protocol (../../framework.md), avoiding external or untrusted network dependencies.
[SAFE]: Tool permissions are limited to Read and AskUserQuestion, which are necessary and sufficient for its stated purpose as a dialogue agent.
[SAFE]: No evidence of prompt injection, data exfiltration attempts, obfuscation, or unauthorized command execution was detected.

marcus-aurelius