The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill utilizes the scan_app tool, which ingests data from external URLs (including health routes, body content, and third-party scripts) to suggest monitoring configurations. This external data ingestion provides a surface for indirect prompt injection, as malicious content on a scanned website could potentially influence the agent's reasoning or the suggestions it provides to the user.
[EXTERNAL_DOWNLOADS]: The skill references and encourages the use of an external MCP endpoint at https://staygreen.io/mcp and a local stdio server available via npm (npx -y staygreen). These resources are provided by the skill's author and are integral to the skill's functionality for interacting with the service.
[COMMAND_EXECUTION]: The documentation provides specific shell commands for the agent or user to execute, including claude mcp add for configuring the MCP server and curl commands for testing heartbeat URLs and triggering failure drills. The skill correctly identifies that heartbeat URLs should be treated as secrets and managed through environment variables.

uptime-monitoring