AI Transparency Statement
Understand how GCGY AI capabilities operate across GrowthOS (AI Corner Coach) and GrowCode (AI Code Review).
GCGY incorporates artificial intelligence to deliver contextual personal coaching (GrowthOS AI Corner Coach) and automated software quality feedback (GrowCode AI Code Review). We believe transparency regarding models, context windows, and limits builds agency and trust.
Primary Inference
OpenAI models (GPT-4 series) power daily habit coaching, mission advice, and GrowCode AI Code Review feedback.
Failover Routing
Google Gemini 1.5 Pro serves as active failover if primary inference endpoints experience latency.
Resiliency Fallback
Anthropic Claude 3.5 Sonnet handles complex multi-day trend reasoning and secondary API failover.
Ecosystem AI Feature Breakdown
GrowthOS AI Corner Coach
Appends 7-day habit heatmaps, active missions, focus domain targets, and recent journal context to deliver personalized habit interventions.
GrowCode AI Code Review
Analyzes submitted code snippets for time complexity (Big-O), spatial efficiency, clean code practices, and potential edge-case bug vulnerabilities.
Cost Guard & Query Caching (`/aiCache`)
To maintain platform performance and prevent unnecessary API cost accumulation, GCGY implements SHA256 prompt hashing. Generic queries are retrieved instantly from the platform's `/aiCache` Firestore collection. Personalized prompts containing state indicators automatically bypass cache lookup for real-time inference.
Server Safety Switches (`NORMAL` / `LIMITED` / `DISABLED`)
All AI API routes evaluate a server-monitored safety kill-switch configuration: