spatialml-debugging

Warn

Audited by Gen Agent Trust Hub on Oct 27, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses adb shell setprop to modify the debug.pico.spatialml.debug system property on a connected device. This represents an impactful command-line operation that changes system state.
  • [PRIVILEGE_ESCALATION]: By setting debug.pico.spatialml.debug to 1, the skill explicitly enables a "device-wide tensor readback permission/AppOps bypass," granting access to data buffers that are typically protected by the operating system's security model.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to capture and analyze untrusted data from the device's system log (logcat).
  • Ingestion points: The spatialml-session.log file populated via adb logcat.
  • Boundary markers: Absent; the skill does not use specific delimiters or instructions to prevent the agent from following directions that might be embedded in the log output.
  • Capability inventory: The agent has capabilities to run adb commands, read/filter files via rg or Select-String, and execute logic based on the findings.
  • Sanitization: No validation or sanitization of the log content is performed before processing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 27, 2026, 07:19 AM