Home

SQLumAI Documentation

Welcome to SQLumAI — an AI‑powered, non‑blocking proxy for Microsoft SQL Server. Use this index to navigate all docs.

Overview

• Architecture: overall system components and data flow.
• MVPs: capabilities across MVP 1–4 and configuration overview.

Enforcement & Policy

• Enforcement Guide: dry‑run vs enforce, rules model, thresholds, env gating.
• QA Checklist: release readiness and manual verification steps.
• Test Strategy: coverage goals, patterns, and CI notes.

How‑To Guides

• Reports: generate daily and dry‑run reports; where artifacts are written.
• Integration: running locally and via Docker, external data sources.
• Examples: small examples of inputs/outputs and typical flows.

Metrics & Monitoring

• Metrics Dashboard: Prometheus endpoint and Grafana setup.
• Dry‑Run Chart: how dry‑run summaries are visualized.

LLM & Insights

• LLM Providers: local Ollama and OpenAI‑compatible endpoints; environment variables.
• Insights: how decisions and profiles become actionable insights.

Data & Internals

• Dry‑Run JSON: on‑disk JSONL formats for decisions and examples.
• RPC Builder: parameter encoding for SQL Server RPC calls.

Operations

• High Availability: proxy topology and failover tips.
• Release Process: versioning and publish steps.
• Release Notes: notable changes by version.

Performance

• Benchmarks: parser and RPC encoding microbenchmarks and methodology.

Tip: For quick commands, Makefile targets are documented in the repository README (setup, dev, test, fmt, lint, clean).