android-camera-bugfix-skill

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external system data including Android logcat outputs and system dumps, which represents a surface for indirect prompt injection.
  • Ingestion points: The skill instructs the agent to collect and analyze data from files such as 'camera_bug.log', 'dumpsys_camera.txt', and system tombstones.
  • Boundary markers: There are no explicit delimiters or boundary markers defined in the instructions for isolating log content.
  • Capability inventory: The skill utilizes 'adb' for device interaction and 'codegraph' for code exploration.
  • Sanitization: The skill provides proactive security guidance in its 'Guardrails' section, instructing users and the agent to redact sensitive data, such as device identifiers and capture payloads, before processing logs.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 05:52 PM
Security Audit — agent-trust-hub — android-camera-bugfix-skill