The Agent Skills Directory

[COMMAND_EXECUTION]: The SKILL.md file provides a feedback command npx --yes submit-expo-feedback@latest --category skills --subject "eas-app-stores" "<actionable feedback>" which interpolates user-provided text directly into a shell command. This allows for potential command injection if the input is not sanitized by the executing agent.
[INDIRECT_PROMPT_INJECTION]: The skill exposes a vulnerability surface where untrusted data (user feedback) is ingested and executed within a shell context.
Ingestion points: The <actionable feedback> placeholder in SKILL.md.
Boundary markers: None. The input is placed directly inside double quotes in a shell command.
Capability inventory: Shell command execution via npx and global CLI tools.
Sanitization: No instructions are provided to sanitize or validate the user-provided feedback string before interpolation.
[DYNAMIC_EXECUTION]: In references/app-store-metadata.md, the skill describes the use of store.config.js to provide dynamic metadata. This file is executed as a Node.js module during the eas metadata:push process, allowing for arbitrary code execution (e.g., fetching data from external URLs) during the deployment pipeline.
[EXTERNAL_DOWNLOADS]: The skill instructs the user or agent to download and execute several external packages using npm and npx, including eas-cli, submit-expo-feedback, and testflight.

eas-app-stores