The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill enables an agent to process, migrate, and review local project files such as .xcstrings, .strings, and .stringsdict. These files, which may contain untrusted content from various contributors, represent an entry point for instructions that could attempt to manipulate the agent's code generation output.
Ingestion points: Reading and extracting keys/values from Localizable.xcstrings and related localization resource files.
Boundary markers: The instructions lack specific delimiters or instructions to ignore potential commands embedded within translation strings.
Capability inventory: The agent is tasked with generating Swift code, recommending build settings, and suggesting CI scripts based on the ingested localization data.
Sanitization: The skill does not implement sanitization or schema validation for the strings found within the localization catalogs before using them in generated code snippets.
[EXTERNAL_DOWNLOADS]: The reference material in references/string-catalogs.md suggests the use of a third-party Swift Package Plugin, xcstrings-tool (hosted at github.com/liamnichols/xcstrings-tool), for managing generated symbols in multi-module environments. This is a recommendation for a community-maintained resource.

ios-localization