The Agent Skills Directory

[SAFE]: The skill is a well-structured set of markdown instructions and rules. It does not contain any executable code, scripts, or commands that could compromise the host system.
[SAFE]: No data exfiltration or credential harvesting patterns were found. The skill does not perform network operations or access sensitive system files like SSH keys or environment secrets.
[SAFE]: While the skill processes untrusted user data (code diffs and documentation), which is a common surface for indirect prompt injection, it handles this by using a blind subagent focused solely on the review task. The ingestion points are located in references/reviewer-prompt.md ({{TARGET_CONTENT_OR_PATHS}} and {{BLUEPRINT_SECTION}}). No explicit boundary markers or sanitization logic are present, which is typical for this type of auditing tool and does not pose a significant security risk.

feature-arch-gate