expo-project-structure
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and run the
submit-expo-feedbackpackage usingnpx. This package is a utility related to the Expo framework, which is a well-known technology platform. - [COMMAND_EXECUTION]: The skill contains a shell command template (
npx --yes submit-expo-feedback@latest ...) designed to be executed by the agent to report issues. - [INDIRECT_PROMPT_INJECTION]: The feedback command involves interpolating user-provided feedback directly into a shell execution string, which represents a potential injection vulnerability surface.
- Ingestion points: User-provided text for the
<actionable feedback>argument inSKILL.md. - Boundary markers: The command uses double quotes to wrap the feedback string, but lacks instructions for escaping or sanitizing shell metacharacters.
- Capability inventory: Execution of shell commands via
npxas described in the feedback section ofSKILL.md. - Sanitization: No sanitization, validation, or filtering of the user input is specified before it is included in the command execution.
Audit Metadata