mobile-submission
Warn
Audited by Gen Agent Trust Hub on Aug 3, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCREDENTIALS_UNSAFECOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install and update third-party CLI tools using Homebrew from repositories that are not recognized as trusted vendors (
brew upgrade asc,brew upgrade tamtom/tap/gplayinreferences/apple.mdandreferences/google.md). - [REMOTE_CODE_EXECUTION]: The instructions direct the agent to fetch and register 'companion skill packs' from external GitHub repositories (
rorkai/app-store-connect-cli-skillsandtamtom/gplay-cli-skills) to extend its functionality at runtime. - [CREDENTIALS_UNSAFE]: The skill operates using high-privilege credentials, including App Store Connect Private Keys and Google Play Service Account JSON files. It explicitly identifies sensitive environment variables (
ASC_PRIVATE_KEY,GOOGLE_APPLICATION_CREDENTIALS) for the agent to use, which increases the risk of credential exposure if the agent's context is compromised. - [COMMAND_EXECUTION]: Multiple shell commands are utilized to perform app builds, uploads, and store management tasks (
asc publish,gplay publish,asc builds upload). - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) as it reads and processes 'rejection notices' and 'reviewer messages' from external store APIs, which are then used to influence agent reasoning and troubleshooting steps.
Audit Metadata