The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources during the destination research phase.
Ingestion points: Destination research, weather patterns, and local insights are gathered via the web_search tool.
Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands within the searched content.
Capability inventory: The skill has access to Bash and write_file tools, which could potentially be targeted by instructions found in external web content.
Sanitization: There are no explicit sanitization or filtering steps defined for the external data before it is incorporated into the itinerary or processed by tools.
[COMMAND_EXECUTION]: The skill is configured with Bash in its allowed-tools list. While the instructions do not currently contain any specific shell commands, the presence of this tool provides a capability that should be monitored if the agent processes untrusted input.

travel-planner