warden-sweep
Warn
Audited by Socket on May 16, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
Mostly aligned with its stated code-review purpose, but it is high-impact automation: it scans all repo content, invokes subagents on untrusted code, writes files, pushes branches, and opens GitHub issues/PRs. This looks more risky than malicious; the main concerns are autonomous repository actions and prompt-injection exposure from full-repo processing, not credential theft or covert exfiltration.
Confidence: 87%Severity: 69%
Audit Metadata