The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted content from English README files to generate translations, creating an indirect prompt injection surface.
Ingestion points: The agent reads the full content of README.md and README_ja.md files during the drift analysis and translation steps.
Boundary markers: There are no instructions to use XML tags, delimiters, or system-level warnings to prevent the agent from obeying instructions embedded within the README files being translated.
Capability inventory: The skill utilizes Read, Write, and Edit tools, allowing an attacker who controls a README file to potentially influence file system modifications.
Sanitization: The skill does not perform any validation or filtering on the ingested text before the two-pass translation process.

sync-ja