gradle-build

Pass

Audited by Gen Agent Trust Hub on Jul 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [SAFE]: No malicious code, obfuscation, persistence mechanisms, or unauthorized data exfiltration patterns were detected in the skill instructions or references.
  • [COMMAND_EXECUTION]: The skill uses standard Gradle wrapper commands (./gradlew assembleDebug, ./gradlew test) for build validation, which is appropriate for its stated purpose.
  • [EXTERNAL_DOWNLOADS]: References to external companion skills from the android and vercel-labs organizations are documented as optional resources for workflow alignment and repository maintenance.
  • [PROMPT_INJECTION]: The skill processes build configuration files, which presents an indirect prompt injection surface. Ingestion points: build.gradle.kts, settings.gradle.kts, and libs.versions.toml. Boundary markers: Absent. Capability inventory: Execution of local Gradle build tasks. Sanitization: Absent. These ingestion points are standard for build management functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 6, 2026, 12:31 PM
Security Audit — agent-trust-hub — gradle-build