android-testing-unit
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user descriptions of Android components to generate and verify tests, creating an attack surface for indirect prompt injection. Ingestion points: User requirements for unit tests (SKILL.md). Boundary markers: Absent; no specific delimiters or instructions to ignore embedded content are provided. Capability inventory: Shell execution of Gradle tasks and Python evaluation scripts (SKILL.md, scripts/run_examples.sh). Sanitization: Absent; no input validation or sanitization is documented for user-provided code specifications.
- [COMMAND_EXECUTION]: The skill executes local shell commands to run unit tests and evaluation triggers. Evidence: Execution of
./gradlew :app:testDebugUnitTestandpython3 scripts/eval_triggers.pyin SKILL.md and scripts/run_examples.sh. - [EXTERNAL_DOWNLOADS]: The skill uses the Gradle wrapper, which automatically fetches the official Gradle distribution if it is not present on the system. Evidence: Usage of
./gradlewin scenarios (references/scenarios.md).
Audit Metadata