What problem does it solve? Static code review can only guess at performance problems, while a live Convex deployment already knows which functions hit read limits, scan too many bytes, or suffer OCC write contention. This Skill reads the deployment's actual 72-hour insight events and root-causes each one in the real code, turning runtime symptoms into evidence-backed, fixable findings. ## Core Features & Use Cases - Runtime Insight Analysis: Reads typed 72h health events (documentsRead/bytesRead limits and thresholds, OCC retries and permanent failures) via the official Convex MCP tools. - Code-Level Root Causing: Traces each insight event to the exact function and line, identifying unindexed filters, missing pagination, or read-modify-write hotspots. - Structured Findings with Fixes: Emits severity-ranked findings on a findings bus with concrete repairs such as adding indexes, using .paginate, or adopting sharded-counter components. - Use Case: Your Convex app's messages:list query reads 4.2MB per call and hits document read limits. The advisor identifies the unindexed .collect() on the messages table, cites the file and line, and proposes an index plus pagination fix. ## Quick Start Ask the advisor to check my Convex deployment's recent insights and root-cause any performance or contention issues it finds.
