The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to install external software using npx -y @mvanhorn/printing-press-library and go install github.com/mvanhorn/printing-press-library/.... While these resources match the vendor's identity, they involve downloading and executing code from external registries and repositories.
[COMMAND_EXECUTION]: The skill relies on executing the hotelist-pp-cli binary with various arguments. It also includes a which command that uses natural language to map user intent to specific CLI commands, which could potentially be manipulated.
[DATA_EXFILTRATION]: The CLI tool provides a --deliver webhook:<url> flag, enabling the agent to send command outputs to any remote HTTP endpoint. Additionally, the feedback --send command provides a mechanism to exfiltrate local feedback data to a configured remote endpoint.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill processes "real traveler reviews" and AI-generated scores scraped from Hotelist.com.
Boundary markers: The instructions do not specify any delimiters or safety markers to differentiate between the tool's output and embedded instructions within traveler reviews.
Capability inventory: The skill has capabilities for file system writes (--deliver file:<path>), network operations (--deliver webhook:<url>), and shell command execution.
Sanitization: The skill does not describe any sanitization or validation of the external hotel data before it is presented to the agent.

pp-hotelist