hmos-local-test

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes the 'hvigorw' tool to execute tests. This command execution is performed through a Python script using list-based arguments with 'subprocess.run', which effectively mitigates the risk of shell injection.
  • [EXTERNAL_DOWNLOADS]: The skill requires the 'json5' Python library to parse project configuration files. This is a standard and legitimate dependency for HarmonyOS development environments.
  • [PROMPT_INJECTION]: The skill instructs the agent to read and interpret project files (such as 'build-profile.json5' and '*.test.ets') to determine the test scope, which represents a surface for indirect prompt injection.
  • Ingestion points: The agent reads configuration and source files from the user-provided project directory.
  • Boundary markers: No specific markers are used to distinguish file content from the agent's instructions.
  • Capability inventory: The skill has the ability to read project files and execute build commands via the provided Python script.
  • Sanitization: The execution script uses safe command construction, ensuring that data extracted from files cannot lead to shell command injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 10:16 AM
Security Audit — agent-trust-hub — hmos-local-test