The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends the installation of official vendor-supported packages, including @tigrisdata/cli via npm and tigris-boto3-ext via pip, to facilitate interaction with Tigris storage services.
[COMMAND_EXECUTION]: The instructions include standard shell commands for deploying assets (e.g., tigris cp) and framework-specific tasks such as rails assets:upload or npm run build. These operations are consistent with the skill's stated purpose of managing asset pipelines.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process local build artifacts for upload to remote storage, creating a potential surface for indirect injection if the build process is compromised.
Ingestion points: The skill reads from local directories such as dist/, .next/static/, and public/assets/ as defined in SKILL.md and framework-specific resource files.
Boundary markers: No explicit delimiters or boundary warnings are used, as the skill targets raw static assets (CSS, JS, fonts).
Capability inventory: The skill utilizes file system read capabilities and network write capabilities through the Tigris CLI and SDKs.
Sanitization: The skill does not perform content sanitization on the assets before deployment, which is standard behavior for static file hosting utilities.

tigris-static-assets