android-apk-audit

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The prompt contains explicit examples that embed secrets on the command line (e.g., apksigner with --ks-pass pass:myPassword) and workflows that search for and report hardcoded secrets, which require the agent to read and potentially emit secret values verbatim in commands and findings.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.98). The skill contains explicit offensive capabilities — ready-to-run Frida scripts (SSL pinning/root/RASP/biometric bypass, keystore inspection, native hooks, memory read/write), APK repackaging/smali patching examples, exploit PoCs (intent injection, SQL/path traversal), ROP/UAF/native exploitation guidance and helper tooling — all of which enable implanting backdoors, disabling protections, exfiltrating secrets and achieving remote or local code execution when misused.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill's Phase 4 dynamic analysis explicitly captures and interprets live app-loaded content (e.g., assets/frida-scripts/network-interceptor.js and webview-monitor.js, plus SKILL.md/README examples showing decrypted/intercepted HTTP requests and WebView payloads) so the agent ingests arbitrary external/untrusted network/WebView content and uses it to drive analysis and PoC generation, which could enable indirect prompt injection.