expo-upgrade

Pass

Audited by Gen Agent Trust Hub on Oct 5, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill triggers the download and execution of packages from external sources during the migration process.
  • Downloads and runs submit-expo-feedback@latest to transmit feedback and context.
  • Downloads and executes the xcobra utility to evaluate code in the application context.
  • Downloads configuration and migration tools from Expo's official registries, such as expo-codemod and expo-doctor.
  • [DYNAMIC_EXECUTION]: The migration process involves evaluating JavaScript code in the running application environment to verify its state.
  • Evidence: Uses bunx xcobra expo eval "_IS_FABRIC" to verify whether the New Architecture rendering system is active.
  • [DATA_EXFILTRATION]: The skill includes a feedback mechanism that transmits user-provided content and project context to an external service.
  • Evidence: The submit-expo-feedback command sends actionable feedback and relevant context to a remote endpoint for analysis.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a large attack surface for indirect prompt injection as it processes and modifies the entire application codebase.
  • Ingestion points: Reads package.json, app.json, and all project source files matching patterns such as **/*.{ts,tsx,js,jsx}.
  • Capability inventory: The skill performs file system cleanup (rm -rf), dependency installation (npx expo install), and code modification via automated codemods.
  • Boundary markers: The instructions do not define delimiters or provide specific warnings to the agent to ignore instructions that might be embedded in the code or configuration files it processes.
  • Sanitization: There is no evidence of content sanitization or validation of the data ingested from the project files before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 5, 2026, 09:27 PM