saizeriya-cli
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill uses
npxandbunxto download and run thesaizeriya.jsandqr-scanner-clipackages from the NPM registry. These tools are required for the skill's stated purpose of restaurant ordering and QR code processing. - [COMMAND_EXECUTION]: The agent executes various shell commands to manage ordering sessions and scan QR codes from image files. Safety rules are included that mandate user confirmation for real-world actions such as submitting orders or calling staff.
- [PROMPT_INJECTION]: The skill has an architectural surface for indirect prompt injection because it processes untrusted data from scanned QR codes. Evidence: 1. Ingestion points: Image files processed in
SKILL.md. 2. Boundary markers: Absent; the scanner output is passed directly as a command argument. 3. Capability inventory: Subprocess execution vianpxandbunx. 4. Sanitization: Relies on natural language instructions for the agent to treat the data as untrusted rather than programmatic filtering.
Audit Metadata