ios-app-store-submission

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to navigate and interact with external web content on App Store Connect, which serves as an untrusted data source.\n
  • Ingestion points: External web content from appstoreconnect.apple.com accessed via browser tools (e.g., tabs_context_mcp).\n
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between the skill's instructions and data retrieved from the website.\n
  • Capability inventory: The agent has access to powerful local tools including xcodebuild, npm, npx, and file system operations across the workspace.\n
  • Sanitization: There is no evidence of sanitization or validation for the data retrieved from the web interface before it influences agent actions.\n- [COMMAND_EXECUTION]: The workflow relies heavily on the execution of shell commands, which is expected for a build tool but represents a broad attack surface if inputs are not carefully managed.\n
  • Evidence: The skill specifies sequences of commands such as xcodebuild -workspace ..., npm install, and npx expo prebuild --clean. These commands use directory names and schemes identified at runtime as arguments.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 07:12 AM
Security Audit — agent-trust-hub — ios-app-store-submission