The Agent Skills Directory

[SAFE]: The skill documentation focuses on legitimate tool integration and operational safety for AI agents. It does not contain any malicious instructions, obfuscation, or unauthorized data access patterns.
[COMMAND_EXECUTION]: The skill provides various examples of gt and git commands. It emphasizes the use of the --no-interactive flag, which is a recognized best practice for preventing AI agents from becoming unresponsive due to interactive shell prompts.
[DATA_EXPOSURE]: The skill describes the location and purpose of Graphite metadata files within the repository's .git directory (e.g., .graphite_cache_persist). This information is used for local stack visualization and does not involve sensitive credential harvesting or data exfiltration to external domains.

erk-gt