ios-localize-copy
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to utilize local command-line tools including
find,xcrun xcstringstool,jq, andgitto synchronize, compile, and validate localization catalogs. These operations are restricted to the local development environment and are standard for the described iOS development tasks. - [INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection as it processes user-facing text from external files for translation.
- Ingestion points: The agent reads data from
.xcstringscatalogs,release_notes.txt, and App Store metadata files (found inSKILL.md). - Boundary markers: Missing (in
SKILL.md). While the instructions mandate structural validation, they do not include explicit prompt delimiters to separate instructions from data. - Capability inventory: File system access and execution of development-related shell commands such as
xcrun,jq, andgit(found inSKILL.md). - Sanitization: Present (in
SKILL.md). Includes placeholder signature comparison and structural diffing.
Audit Metadata