The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires installing external software via npx -y @mvanhorn/printing-press and go install github.com/mvanhorn/printing-press-library/.... While these are vendor-provided resources, they originate from an author not on the trusted vendors list.
[DATA_EXFILTRATION]: The CLI tool supports a --deliver webhook:<url> flag, which allows redirecting command output (including travel availability and trip revalidation details) to arbitrary remote endpoints. This creates a risk of data exfiltration if the agent is directed to use a malicious URL.
[COMMAND_EXECUTION]: The skill executes the seats-aero-pp-cli binary through shell commands. While used for legitimate search tasks, arbitrary argument injection could pose a risk if not properly handled by the underlying tool.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes data from external APIs and possesses high-privilege capabilities such as file system and network write access.
Ingestion points: Data is ingested from the Seats.aero Partner API through the availability, routes, and trips commands (SKILL.md).
Boundary markers: No delimiters are specified to protect the agent context from potentially malicious instructions embedded in the API responses.
Capability inventory: Shell command execution, network POST requests (--deliver webhook:<url>), and local file writes (--deliver file:<path>).
Sanitization: There is no documented validation or sanitization of data returned from the API before it is processed by the agent.

pp-seats-aero