expo-dev-client
Warn
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to use
npx --yes submit-expo-feedback@latest. This command downloads and executes unversioned code from a third-party package on the NPM registry at runtime, which poses a supply chain risk if the package is compromised. - [COMMAND_EXECUTION]: The skill provides instructions to run several shell commands including
eas build,adb install,xcrun, andideviceinstaller. While these are standard for mobile development, they grant the agent significant interaction with the host system and cloud credentials. - [INDIRECT_PROMPT_INJECTION]: The skill processes project configuration files and build service outputs, creating an attack surface where malicious data could influence agent behavior.
- Ingestion points: Reads project configuration from
eas.jsonand monitors output/logs fromeas buildcloud or local processes. - Boundary markers: The instructions do not define delimiters or provide specific warnings to the agent to disregard instructions embedded within these data sources.
- Capability inventory: The skill has access to powerful shell commands via the mobile toolchain and
npx. - Sanitization: No explicit validation or sanitization logic is provided for inputs derived from local files or build services.
Audit Metadata