The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests and processes data from the App Store Connect web interface.
Ingestion points: The agent reads UI elements, labels, and asynchronous loading states (e.g., "Loading...") from the appstoreconnect.apple.com domain.
Boundary markers: No explicit delimiters or boundary markers are defined to separate instructions from the web data being processed.
Capability inventory: The skill utilizes browser automation (clicking, filling forms, scrolling) and executes system commands using the asc CLI tool.
Sanitization: Input parameters such as the app name, SKU, and bundle ID are interpolated into browser actions and CLI commands without evidence of sanitization or validation.
[COMMAND_EXECUTION]: The skill relies on the execution of multiple shell commands via a CLI tool named asc to interact with App Store Connect APIs.
Evidence: The workflow involves commands like asc bundle-ids create, asc apps list, asc app-setup info set, and asc pricing availability create across several steps.
Context: These commands are used to manage App Store Connect resources and are consistent with the skill's stated purpose of automating app creation.

asc-app-create-ui