The Agent Skills Directory

[SAFE]: The skill provides instructions for using the forge comment command to document progress on tasks. It follows standard documentation practices without introducing security vulnerabilities.
[NO_CODE]: The skill consists entirely of markdown instructions and command examples. No external scripts, binaries, or configuration files are executed or downloaded.
[INDIRECT_PROMPT_INJECTION]: The skill mentions that comments are visible to other agents, creating a communication channel between sessions. While this is an attack surface for indirect prompt injection, it is the intended primary purpose of a collaboration tool and contains no specific exploitable patterns in the instructions themselves.

comment