anolisa-register

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEPRIVILEGE_ESCALATIONDATA_EXFILTRATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [PRIVILEGE_ESCALATION]: The skill utilizes sudo for registration and unregistration commands, which are required for system-level configuration changes.
  • [DATA_EXFILTRATION]: The skill manages the transmission of anonymized telemetry data to the vendor's infrastructure. It provides a clear unregistration mechanism to stop data collection.
  • [COMMAND_EXECUTION]: The skill executes the anolisa CLI utility, which is the management tool for the system.
  • [INDIRECT_PROMPT_INJECTION]: The skill maps natural language intents directly to administrative commands, creating a surface for indirect influence. * Ingestion points: Natural language intent mappings in SKILL.md. * Boundary markers: None explicitly defined in the skill mappings. * Capability inventory: Execution of sudo commands and interaction with the anolisa CLI in SKILL.md. * Sanitization: Relies on predefined command structures and human-in-the-loop confirmation for registration actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 03:58 AM