app-store-release-kit
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes local source code and presents excerpts to the agent, which could contain malicious instructions designed to hijack the agent's task.
- Ingestion points: The script
scripts/audit_app_store.pyiterates through and reads content from various source files (including .swift, .js, .ts, .m, and .plist) within a local directory specified by the user. - Boundary markers: Audit findings include raw 180-character code snippets provided in a Markdown report. There are no delimiters or specific instructions provided to the agent to disregard or treat as data any potential commands found within these snippets.
- Capability inventory: The skill enables the agent to run shell commands (the provided Python scripts), write to the local file system (via
init_release_assets.py), and read local project files. - Sanitization: Content excerpts are truncated by the script but do not undergo sanitization or instruction filtering before being displayed to the agent in the final report.
Audit Metadata