The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the internet via search tools, which could potentially be manipulated by malicious web content to influence the agent's behavior.
Ingestion points: The skill uses internet_search results to identify attractions, local guides, and restaurant recommendations.
Boundary markers: No specific delimiters or instructions to ignore embedded commands within search results are defined in the SKILL.md instructions.
Capability inventory: The agent has access to specific travel tools for ticket queries (mcp_12306_query-tickets) and location services (mcp_amap-amap-sse_maps_*), which are low-impact capabilities.
Sanitization: There is no explicit sanitization or filtering of the content fetched from the web before it is incorporated into the final travel guide output.

travel-guide