android-additional-tests
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- Indirect Prompt Injection (LOW): The skill performs direct string interpolation of user-provided inputs into generated Kotlin code and YAML configurations. Ingestion points:
package_nameandmain_activityinputs in the YAML frontmatter. Boundary markers: None. Capability inventory: The skill generates and writes multiple Kotlin files to the local filesystem and provides commands to execute them via the Gradle wrapper (./gradlew). Sanitization: Absent. There is no escaping or validation of these strings before they are injected into the source code templates, potentially allowing for code injection if malicious values are provided as inputs. - External Downloads (SAFE): The GitHub Actions workflow references several external actions including
reactivecircus/android-emulator-runnerandandroid-actions/setup-android. While these are not hosted by the specific organizations in the trusted sources list, they are widely recognized community standards for Android CI/CD pipelines and are used here in a conventional manner.
Audit Metadata