The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill setup process involves downloading the tesla-http-proxy tool from the official Tesla Motors GitHub repository (github.com/teslamotors/vehicle-command) to handle vehicle command signing.
[COMMAND_EXECUTION]: The start_proxy.sh script executes the downloaded tesla-http-proxy binary in the background to act as a local signing service.
[COMMAND_EXECUTION]: tesla_fleet.py uses os.execv to dynamically route user commands to specific sub-scripts (e.g., command.py, vehicle_data.py) within the skill directory.
[PROMPT_INJECTION]: The skill architecture includes an indirect prompt injection surface (Category 8).
Ingestion points: vehicle_data.py retrieves and displays data from the Tesla Fleet API, including vehicle display names and configuration metadata.
Boundary markers: Output formatting does not include clear delimiters or instructions for the agent to treat API-retrieved strings as untrusted data.
Capability inventory: The skill possesses significant capabilities, including the ability to unlock doors, honk the horn, and control vehicle charging/climate via command.py.
Sanitization: Content returned from the API is parsed and printed without additional sanitization or verification, which could allow maliciously crafted vehicle metadata to influence agent behavior.

tesla-fleet-api