apk-redteam-pipeline

Pass

Audited by Gen Agent Trust Hub on Nov 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches application packages from third-party Android repositories including apkpure.net and apkmirror.com to facilitate security analysis. It also provides instructions for downloading the jadx decompilation tool from GitHub.
  • [COMMAND_EXECUTION]: The workflow utilizes a wide range of command-line utilities for data processing, extraction, and network inspection, including curl, grep, 7z, openssl, and the Android Debug Bridge (adb).
  • [INDIRECT_PROMPT_INJECTION]: Because the skill is designed to process and analyze untrusted third-party mobile applications, it possesses a significant attack surface for indirect instruction injection.
  • Ingestion points: Analysis targets include APK binaries, resource files (strings.xml, AndroidManifest.xml), and external responses from Firebase or other cloud service endpoints.
  • Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard potentially malicious instructions embedded within the extracted strings or metadata of the applications being analyzed.
  • Capability inventory: The skill environment supports comprehensive shell execution, file system access, and network communication capabilities, which are required for mobile application research.
  • Sanitization: The skill performs pattern matching (grep) but does not implement sanitization or validation of the extracted data before displaying it or using it in further analytical steps.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 12, 2026, 05:47 PM