gplay-release-flow
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read release notes and metadata from external files and directories. This establishes a potential indirect prompt injection surface where content from these files could influence agent behavior.
- Ingestion points: File paths passed to the
--release-notesflag and directory paths passed to--listings-dirand--screenshots-dirinSKILL.md. - Boundary markers: There are no instructions providing specific delimiters or 'ignore' commands for the content of these external files.
- Capability inventory: Shell command execution via the
gplayCLI for app uploads, track promotion, and rollout management. - Sanitization: No explicit instructions are provided for validating or sanitizing the content of the external files before processing.
Audit Metadata