build-iphone-apps

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHCOMMAND_EXECUTIONCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADS
Full Analysis
  • COMMAND_EXECUTION (HIGH): The files references/cli-workflow.md and references/ci-cd.md contain instructions to use sudo xcode-select. Granting an AI agent the ability to execute sudo commands for system-level modifications presents a high risk of privilege escalation and host compromise.
  • CREDENTIALS_UNSAFE (HIGH): In references/app-store.md, hardcoded credentials (demo@example.com / Demo123!) are provided. While intended as a template for App Store reviewers, hardcoding any username/password combination in source or reference files is a high-risk security violation.
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill relies on multiple external dependencies. references/cli-workflow.md and references/ci-cd.md recommend brew install for tools like xcodegen, fastlane, ios-deploy, and swiftlint. These represent unverified third-party binaries that are downloaded and executed at runtime.
  • COMMAND_EXECUTION (MEDIUM): SKILL.md includes a command piping output to xcsift. This tool is not a standard part of the iOS/Xcode development suite and is not installed by the provided setup scripts, making it an unverifiable and potentially malicious dependency.
  • DATA_EXFILTRATION (LOW): The skill workflows in references/ci-cd.md and references/cli-workflow.md handle sensitive data including .p12 certificates, .p8 API keys, and provisioning profiles. While typical for CI/CD, processing these artifacts via agent scripts increases the surface area for credential exposure.
  • PROMPT_INJECTION (LOW): The skill is susceptible to indirect prompt injection.
  • Ingestion points: User descriptions of app logic are used to generate Swift source code in workflows/build-new-app.md.
  • Boundary markers: Delimiters are absent when interpolating user requirements into generated files.
  • Capability inventory: The skill has access to powerful CLI tools like xcodebuild and xcrun in SKILL.md.
  • Sanitization: There is no evidence of sanitization or validation of user-provided strings before they are written to code files and executed.
Recommendations
  • AI detected serious security threats
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 05:45 PM