The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external search results which could be used to influence agent behavior through malicious item or restaurant names.
Ingestion points: The skill processes data from search_restaurants, search_menu, and get_restaurant_menu as seen in SKILL.md.
Boundary markers: The instructions do not specify the use of delimiters or specific warnings to ignore instructions that might be embedded in the menu or restaurant data.
Capability inventory: The agent has access to the place_food_order tool, which executes real-world financial transactions.
Sanitization: No sanitization or validation of the external API content is defined before the agent proceeds to build a cart or place an order.

swiggy-mcp-food