The Agent Skills Directory

[SAFE]: The skill uses local discovery tools like glob and grep to verify file paths and line numbers, ensuring the accuracy of documentation anchors without attempting unauthorized data access.
[SAFE]: Filesystem interactions are restricted to writing structured JSON files within a specific .tours/ directory, and the skill explicitly forbids the modification of existing source code.
[SAFE]: The skill does not include any external dependencies, network operations, obfuscated code, or persistence mechanisms.
[SAFE]: While the skill processes repository content (a data ingestion surface), this is essential for its primary documentation purpose and is managed within a narrow, structured output format.

code-tour