reason_grok

What problem does it solve?

Orchestrates internal Reflexion-style reasoning with external multi-LLM breadth to deliver high-quality, explainable answers.

Core Features & Use Cases

• Structured hybrid reasoning: combines deep internal critique with parallel external evaluation for robust results.
• Evidence labeling & routing: attaches Verified/Inferred/Unproven labels to claims and routes tasks by capability gaps.
• Privacy-aware orchestration: classifies data sensitivity and applies redaction or local handling as needed.
• Use Case: Complex research, architecture decisions, and policy reviews that benefit from multi-model debate and traceable outputs.

Quick Start

Invoke /reason with your question to get a complete, audit-trail answer.