The Agent Skills Directory

[DYNAMIC_CONTEXT_INJECTION]: The skill uses the !command syntax to execute a shell script automatically when the skill is loaded. The command !bash "${CLAUDE_SKILL_DIR}/../../hooks/yaml-helper.sh" resolve_config --keys automation leverages environment variables and relative paths to execute code at load time.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project data including source code, Game Design Documents, and narrative lore. Ingestion points: Reads files for string extraction and translator briefing generation. Boundary markers: Lacks explicit delimiters for content read from files. Capability inventory: Access to Write, Edit, Bash, and Agent tools, allowing for potential modification of files or command execution based on interpreted untrusted content. Sanitization: No evidence of input validation or escaping for external content is provided in the instructions.

localize