raspberry-pi-ai
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPRIVILEGE_ESCALATIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches installation scripts from official service domains (ollama.com) and source code from established repositories (github.com/raspberrypi).
- [REMOTE_CODE_EXECUTION]: Provides procedures for executing remote shell scripts for software deployment, specifically for the Ollama inference engine.
- [PRIVILEGE_ESCALATION]: Includes administrative commands using sudo for managing system services, updating firmware, and configuring network interfaces.
- [PERSISTENCE]: Provides templates and instructions for configuring systemd service units to enable automatic application startup at boot time.
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from network discovery tools and local API endpoints. 1. Ingestion points: Reads network service advertisements via avahi-browse (debugging-playbook.md) and local API outputs from Ollama and Hailo (SKILL.md). 2. Boundary markers: Uses structured checklists and Pre-Flight Checks to delineate setup phases. 3. Capability inventory: Full system access including apt, systemctl, chmod, and sudo across documentation and templates. 4. Sanitization: Emphasizes manual command verification and provides warnings regarding the review of diagnostic reports before sharing.
Audit Metadata