restaurant-search

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to run the hpp CLI tool using the Bash tool. It provides templates that incorporate user-supplied input (like location names and keywords) as arguments for search and lookup operations.
  • [EXTERNAL_DOWNLOADS]: The skill references the author's official GitHub repository (github.com/jackchuka/hpp) as the source for the necessary hpp tool.
  • [CREDENTIALS_UNSAFE]: The skill requires a HOTPEPPER_API_KEY. It correctly identifies that this should be handled as an environment variable, which is the standard secure practice for managing API credentials in this environment.
  • [PROMPT_INJECTION]: Indirect injection surface: The skill processes data from the HotPepper Gourmet API.
  • Ingestion points: Restaurant results and area codes retrieved via hpp search, hpp area, and hpp genre commands.
  • Boundary markers: Absent; external API data is formatted directly into the agent's response to the user.
  • Capability inventory: The skill has access to the Bash tool for running CLI commands.
  • Sanitization: Absent; the instructions do not specify any validation or sanitization for the text returned by the external API before it is presented or processed.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 06:03 PM