app-store-release
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill heavily utilizes shell commands, specifically within
mac_release.sh, to perform tasks such as creating temporary keychains, importing certificates, and executingxcodebuildfor app archiving and exporting. These are standard operations for iOS build automation. - [EXTERNAL_DOWNLOADS]: The GitHub Actions template (
templates/ios-release.yml) fetches scripts from the author's repositorydeonmenezes/mobile-app-skillat a pinned version. This download is used to provide the necessary tooling for cloud-based releases and originates from a documented vendor resource. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process status reports and rejection messages from App Store Connect, which could theoretically contain instructions intended to influence the agent.
- Ingestion points: API response data processed by
asc_status.pyand manual retrieval of reviewer messages described inreferences/rejections.md. - Boundary markers: The instructions do not specify prompt delimiters or sanitization for these external messages, though the agent is directed to report them verbatim to the user.
- Capability inventory: The skill possesses shell execution capabilities (SSH,
xcodebuild) and file writing capabilities (release_config.py,manual_sign.py). - Sanitization: The scripts do not perform input validation or escaping on the strings retrieved from the App Store Connect API.
Audit Metadata