The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from local configuration files while possessing execution capabilities.
Ingestion points: The skill uses read_file, edit_file, and search_code to ingest content from external project configuration files such as eas.json, app.json, and PrivacyInfo.xcprivacy (SKILL.md, references/eas-build-ota-and-store-compliance-specs.md).
Boundary markers: The instructions do not specify the use of delimiters (like XML tags or markdown blocks) or "ignore embedded instructions" warnings when the agent reads or interpolates data from these project files.
Capability inventory: The skill utilizes execute_command for EAS CLI operations (e.g., eas build, eas credentials) and performs file system operations including write_file and create_file (SKILL.md).
Sanitization: There is no evidence of explicit sanitization or validation logic to ensure that content read from project files is not interpreted as instructions by the agent before being passed to CLI tools or written back to the disk.

deploy-mobile-app