android-auto-release

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill generates a GitHub Actions workflow that executes several shell commands to manage the build environment.
  • Decodes a Base64-encoded keystore from secrets to a file for APK/AAB signing.
  • Dynamically creates a local.properties file containing build-time secrets and environment variables.
  • Executes ./gradlew to perform the application build process.
  • [EXTERNAL_DOWNLOADS]: The workflow relies on external GitHub Actions to perform its tasks.
  • Fetches and uses actions/checkout@v4 and actions/setup-java@v4 (official GitHub/well-known actions).
  • Utilizes r0adkll/upload-google-play@v1, which is a widely used third-party action for interacting with the Google Play Developer API.
  • [INDIRECT_PROMPT_INJECTION]: The skill includes a step that parses an external file to generate metadata for a production release.
  • Ingestion points: Reads content from CHANGELOG.md within the repository.
  • Boundary markers: Uses awk to extract specific bulleted lines from the top entry and head to truncate the output.
  • Capability inventory: The pipeline has the capability to write to the file system and upload data to the Google Play Console via the Developer API.
  • Sanitization: The skill applies a length limit (480 characters) to prevent oversized payloads, though it does not explicitly sanitize the text content for malicious instructions directed at the agent or Play Store reviewers.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 07:30 AM
Security Audit — agent-trust-hub — android-auto-release