The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected. The skill uses standard tools for its intended purpose of code review and demonstrates a commitment to security by including examples of identifying and removing hardcoded secrets.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it processes untrusted data from GitLab, such as merge request descriptions, commit messages, and code diffs. This is an inherent and expected risk for a code review tool.
Ingestion points: Untrusted content enters the context via get_merge_request, get_merge_request_commits, and get_merge_request_diffs tools.
Boundary markers: None are explicitly defined in the instructions to differentiate untrusted data from agent instructions.
Capability inventory: The skill relies on read-only GitLab metadata tools; it does not request capabilities for command execution, file-system modifications, or network exfiltration.
Sanitization: No specific sanitization or filtering logic is provided for the ingested content.

review-mr