The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the user or agent to run node .gitnexus/run.cjs analyze in the terminal to refresh the analysis index. This executes a local script within the .gitnexus directory, which is a standard pattern for project-specific tooling.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it analyzes and modifies user-provided source code. Maliciously crafted comments or string literals in the target code could attempt to influence the agent's behavior. The skill mitigates this by recommending dry_run modes and using detect_changes to verify the scope of edits.
[SAFE]: No evidence of data exfiltration, obfuscation, or hardcoded credentials was found. The functionality is consistent with its stated purpose of assisting with code restructuring.

gitnexus-refactoring