The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill is designed to analyze external project files via Read, Glob, and Grep tools. This creates an ingestion surface for untrusted data. If a project file contains malicious instructions within strings or comments, it could potentially influence the agent's reasoning. However, the lack of Write or Shell capabilities in the allowed-tools metadata limits the impact of such an attack.
[Metadata Poisoning] (LOW): There is a minor discrepancy between the metadata and the content. The allowed-tools field specifies Read, Glob, and Grep, but the Script section suggests executing python scripts/i18n_checker.py. This inconsistency could lead to confusion regarding the skill's operational boundaries.
[Unverifiable Dependencies] (LOW): The skill references a local file scripts/i18n_checker.py which was not included in the provided context for analysis. While local script references are common, the behavior of this script cannot be verified.

i18n-localization