gourmet-research

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill retrieves data from various external food review and mapping services, including Google Maps, Tabelog, Retty, and OpenRice, to populate gourmet research files.
  • [COMMAND_EXECUTION]: The instructions direct the agent to utilize Playwright for browser automation when static scraping is insufficient, allowing for the execution of browser-based data extraction tasks.
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface:
  • Ingestion points: External food review websites and mapping platforms provide the primary data source for the skill's workflow (documented in SKILL.md).
  • Boundary markers: The skill lacks explicit instructions to use delimiters or ignore potentially malicious instructions embedded within the scraped external content.
  • Capability inventory: The skill involves writing to the local file system (creating gourmet/ directories and markdown files) and using browser automation tools (Playwright).
  • Sanitization: No specific sanitization or validation steps are defined for the data extracted from external sources before it is processed and stored.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 02:18 PM
Security Audit — agent-trust-hub — gourmet-research