What problem does it solve?
Writing Convex backend code requires strict conventions—object-form functions, validators, index-based reads, and correct generated imports—that generic AI assistants routinely get wrong, causing deploy failures and performance defects.
Core Features & Use Cases
- Correct Function Authoring: Enforces object-form syntax with args and returns validators on every query, mutation, and action.
- Index-First Data Access: Replaces unbounded .collect() and .filter() scans with .withIndex() queries and pagination for scalable reads.
- Deploy Verification: Runs npx tsc --noEmit and convex dev pushes to catch validation and import errors before finishing.
- Use Case: When adding a new messages table with a by-channel query, the skill defines the schema index, writes an internalQuery using withIndex and pagination, and verifies the deployment compiles cleanly.
Quick Start
Ask the agent to add a new Convex query with proper indexes and validators for a table in your convex/ directory.