The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the linq-pp-cli and linq-pp-mcp tools from external repositories.
Evidence: Installation instructions include npx -y @mvanhorn/printing-press-library and go install from github.com/mvanhorn/printing-press-library. These are vendor-provided resources associated with the skill author.
[COMMAND_EXECUTION]: The skill operates by executing shell commands via the linq-pp-cli binary to interact with the Linq Partner API.
Evidence: The instructions provide extensive examples of CLI invocations for messaging (compose), attachment management (attachments), and webhook configuration (webhooks).
[DATA_EXFILTRATION]: The CLI tool includes built-in functionality that enables data transmission to remote endpoints.
Evidence: The --deliver webhook:<url> flag allows the output of any command to be POSTed to a user-specified URL. Additionally, the feedback command is designed to send local data to a configured LINQ_FEEDBACK_ENDPOINT.
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing untrusted external data.
Ingestion points: The skill handles external message text, media URLs, and file attachments through arguments like --text, --url, and --file across multiple commands.
Boundary markers: The instructions do not specify the use of delimiters or warnings to prevent the agent from obeying instructions embedded within the processed data.
Capability inventory: The skill can execute local commands (linq-pp-cli), perform network operations (API calls and webhooks), and write to the local file system using the --deliver flag.
Sanitization: No explicit input validation or sanitization routines are described in the instructions for the content being processed.

pp-linq