android-compose-state-effects

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions and example scenarios involve executing standard development tools. Specifically, it references running ./gradlew for Android testing and assembly, and a local Python script scripts/eval_triggers.py for skill validation.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it is designed to ingest and process user queries and potentially code related to Android Compose.
  • Ingestion points: User-provided problem descriptions or code snippets related to Compose state and effects (SKILL.md).
  • Boundary markers: None explicitly defined in the instructions for separating user data from system instructions.
  • Capability inventory: Execution of local shell scripts and build tools (gradlew) as described in SKILL.md and scenarios.md.
  • Sanitization: No specific sanitization or validation of the input content is described within the skill logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 05:30 PM
Security Audit — agent-trust-hub — android-compose-state-effects