espresso-suite
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data sources such as logcat logs, view hierarchy dumps, and diagnostic reports to guide the agent in fixing and authoring UI tests. This creates an indirect prompt injection surface where malicious content within the logs could potentially influence the agent's code generation and file modifications. Ingestion points: External report files and log outputs (e.g.,
reports/bilal-notes.md,reports/nightly-4712.txt). Boundary markers: None identified; the skill does not specify markers to isolate log content. Capability inventory: Modification of production code, test classes, and build files, alongside shell command execution. Sanitization: None identified; external data is processed directly. - [COMMAND_EXECUTION]: The skill executes
./gradlew connectedAndroidTestto perform instrumentation tests. - [EXTERNAL_DOWNLOADS]: The skill manages standard Android dependencies via Gradle and references the
reactivecircus/android-emulator-runnerGitHub Action for CI workflows.
Audit Metadata