The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructions and metadata guide the installation of CLI tools from the vendor's GitHub repository and NPM organization. Evidence includes go install github.com/mvanhorn/printing-press-library/library/other/apartments/cmd/apartments-pp-cli@latest and npx -y @mvanhorn/printing-press.- [DATA_EXFILTRATION]: The CLI includes built-in features that allow for the transmission of data to external endpoints. The --deliver webhook:<url> flag enables the delivery of command output to arbitrary HTTP destinations, and the feedback command can POST local data to a remote server defined by the APARTMENTS_FEEDBACK_ENDPOINT environment variable.- [PROMPT_INJECTION]: The skill establishes an attack surface for indirect prompt injection by ingesting and processing untrusted HTML content from the web.
Ingestion points: The rentals and listing commands fetch and parse remote HTML from Apartments.com.
Boundary markers: No specific delimiters or instructions are provided to the agent to ignore potentially malicious instructions embedded in the scraped content.
Capability inventory: The agent can execute bash commands and perform network operations via the CLI.
Sanitization: No sanitization or filtering of the extracted web content is documented.

pp-apartments