android-apk-patch

Fail

Audited by Socket on May 6, 2026

3 alerts found:

AnomalySecurityMalware
AnomalyLOW
references/end-to-end-workflow.md

No embedded malware or data-theft logic is present because the fragment is a procedural APK modification/testing workflow rather than executable dependency code. However, it explicitly enables repackaging and includes guidance for bypassing app integrity/signature checks using Frida and then patching smali—capabilities that materially increase misuse potential (tampering/evasion). Treat as a security-relevant tampering toolkit rather than a benign library component.

Confidence: 70%Severity: 62%
SecurityMEDIUM
SKILL.md

SUSPICIOUS: the skill is internally coherent as an APK patching/reversing guide, but it gives an AI agent strong offensive security capabilities: bypassing signatures/integrity checks, rooting environments, dynamic instrumentation, and modifying/distributing patched apps. Install sources are mostly official, lowering malware confidence, yet the overall security risk remains high because the skill meaningfully expands offensive capability and privileged execution scope.

Confidence: 89%Severity: 82%
MalwareHIGH
references/advanced-patching.md

This fragment is a high-risk evasion toolkit: it uses Frida to bypass TLS pinning and fully disable hostname/certificate verification (permissive TrustManager + always-true HostnameVerifier), evades root detection by falsifying filesystem/package presence signals, and forges signature data by injecting attacker-controlled Signature objects. The inclusion of additional integrity/Play Integrity bypass guidance further supports malicious intent. Treat as a severe supply-chain security red flag if distributed or depended upon.

Confidence: 90%Severity: 95%
Audit Metadata
Analyzed At
May 6, 2026, 08:59 AM
Package URL
pkg:socket/skills-sh/CodeAtCode%2Foss-ai-skills%2Fandroid-apk-patch%2F@cecdd37d6e9b4874f72b4a0cddaf24745f551641
Security Audit — socket — android-apk-patch