skills/kazukinagata/shinkoku/furusato/Gen Agent Trust Hub

furusato

Pass

Audited by Gen Agent Trust Hub on Feb 22, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (LOW): The skill ingests untrusted data from external receipt images which are processed via an OCR sub-skill.
  • Ingestion points: Image files listed via Glob and processed by the /reading-receipt skill (SKILL.md Step 1-2).
  • Boundary markers: The skill relies on a structured output block (---FURUSATO_RECEIPT_DATA---) to parse data (SKILL.md Step 1-2).
  • Capability inventory: Executes local CLI commands shinkoku furusato add and shinkoku tax furusato-limit (SKILL.md Steps 2 and 4).
  • Sanitization: Implements a robust 'User-in-the-loop' verification step (Step 1-3) where extracted data must be confirmed or corrected by the user before being passed to executable tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 22, 2026, 11:14 PM