The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts scripts/generate_final_guide.py, scripts/generate_travel_web.py, and scripts/generate_travel_doc.py use subprocess.run() with shell=True to execute the start command. This is used to automatically open the generated itinerary files (HTML or Word) in the user's default browser or document viewer on Windows systems.
[EXTERNAL_DOWNLOADS]: The SKILL.md file instructs the agent to use the WebSearch tool to retrieve live data regarding train schedules, hotel availability, and local attractions. This is a core functional requirement of the travel assistant.
[PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by processing external data from web searches and user inputs.
Ingestion points: User-provided destination, budget, and travel preferences via AskUserQuestion, and live data retrieved via WebSearch (as specified in SKILL.md).
Boundary markers: Absent. The skill does not explicitly instruct the agent to ignore instructions embedded in the search results.
Capability inventory: The skill has file-write capabilities (open("w") in several scripts) and shell execution capabilities (subprocess.run with shell=True).
Sanitization: The scripts implement HTML entity escaping through esc() functions to mitigate the risk of XSS or injection in the final generated travel guide documents.

travel-agent