The Agent Skills Directory

[COMMAND_EXECUTION]: The skill incorporates the user-provided $ARGUMENTS directly into the shell command git flow hotfix start $ARGUMENTS. This creates a command injection vulnerability where a user could provide a version string containing shell metacharacters (e.g., ; rm -rf /) to execute unauthorized commands on the host system.
[PROMPT_INJECTION]: The user-supplied version string is interpolated into a prompt template used to initialize a new agent. This is a prompt injection surface where a malicious user could provide input designed to override the secondary agent's instructions, potentially leading to unauthorized actions or data access.
[DATA_EXFILTRATION]: The skill automatically executes git push -u origin hotfix/$ARGUMENTS. While this is standard for hotfix workflows, the automation of pushing to a remote repository with a user-controlled branch name could be leveraged for data exfiltration if the user has been social-engineered into adding a malicious remote repository.

start-hotfix