orca-onboard
Warn
Audited by Socket on Oct 5, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The core repo-analysis behavior fits the stated purpose, but the skill's footprint is broader than onboarding: it installs another skill, bootstraps harness code from GitHub/raw scripts, modifies user/global Claude setup, starts a local server, and even commits/pushes SKILL changes to a remote repository. These are disproportionate to codebase onboarding and create medium-high supply-chain and execution-trust risk, though there is no clear evidence of credential theft or deliberate exfiltration.
Confidence: 85%Severity: 68%
Audit Metadata