The Agent Skills Directory

[Command Execution]: The skill utilizes the git command-line tool to identify code changes via git diff. This is a standard operation for a code review tool within a development environment.
[Indirect Prompt Injection Surface]: The skill processes user-supplied input through the $ARGUMENTS variable to focus its review agents. This creates a potential surface where untrusted data could influence agent behavior.
Ingestion points: User input is ingested via the $ARGUMENTS variable referenced in SKILL.md.
Boundary markers: There are no explicit delimiters or boundary markers surrounding the interpolation of $ARGUMENTS to distinguish instructions from data.
Capability inventory: The skill has the capability to read local files via git diff, delegate tasks to other agents, and perform file modifications as part of the fixing phase in SKILL.md.
Sanitization: The skill does not implement specific sanitization or validation logic for the content provided in $ARGUMENTS.

code-review