The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data through a user interview (project name, purpose, tech stack) and external project templates, which are then interpolated into documentation skeletons. While {placeholder} markers are used as boundaries, the skill lacks explicit sanitization or escaping of this content before it is written to files that the agent will later read as source-of-truth instructions. The skill also has significant capabilities including file writing and shell command execution.
[COMMAND_EXECUTION]: The skill utilizes shell commands for its core functionality. It executes a version check script using curl and find, uses the GitHub CLI (gh) for template verification, and runs a local Node.js script (node scripts/todos/cli.mjs) for linting project tasks.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch versioning information from https://foundry.thehorizonfoundry.com/api/version and retrieves documentation specifications from the author's public GitHub repository (raw.githubusercontent.com/horizon-foundry/foundry). These operations target the vendor's own infrastructure and are used for maintenance and configuration.

scaffold