r8-analyzer

Pass

Audited by Gen Agent Trust Hub on Oct 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes ./gradlew to generate R8 configuration reports and runs local Python scripts to analyze the results. This is required for the skill's primary function of optimizing Android build configurations.
  • [DYNAMIC_EXECUTION]: The skill provides Python source code within its reference documentation for converting and analyzing Protobuf data, instructing the agent to execute these snippets during the analysis process.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted project data including build.gradle and generated binary R8 reports. While it lacks explicit boundary markers for this data, the risk is low as the data is used for metric extraction and the primary capabilities are restricted to local analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 5, 2026, 09:27 PM