The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were identified in the skill. The technical instructions, commands, and file interactions are consistent with legitimate React Native development workflows.
[INDIRECT_PROMPT_INJECTION]: The skill performs stack detection by reading local configuration files like package.json, app.json, and eas.json (documented in SKILL.md). This is a standard and necessary behavior for providing context-aware development advice and represents no significant security risk.
Ingestion points: Reads package.json, app.json, and eas.json via shell commands (cat, ls).
Boundary markers: Not explicitly defined for the data read from configuration files.
Capability inventory: The skill uses standard development tools including node, npx, pod, eas-cli, and gradlew.
Sanitization: Content from configuration files is used to determine project versions and frameworks but is not directly interpolated into executable commands in a way that suggests vulnerability to injection.

react-native