Architecture quality for AI-driven development
Open-source linters that control quality, cost, and safety of AI agent workflows. All Go, no LLM, under 10ms.
The Pipeline
prompt -> seclint (safe?) -> promptlint (route) -> agent -> archlint (quality) -> costlint (cost)Products
archlint
Quality Gate
Structural graphs, SOLID violations, cycle detection. Scans Go and Rust projects.
GitHubpromptlint
Model Router
Score prompt complexity, route to the cheapest sufficient model. Save 40-60% on AI costs.
GitHubcostlint
Cost Optimizer
Token cost analysis, cache metrics, A/B testing between models. Find where money goes.
GitHubseclint
Content Filter
Age-appropriate ratings (6+/12+/16+/18+). Filter prompts before they reach any model.
GitHubServices
Architecture Audit
Automated scan of your codebase with expert analysis. Structural violations, dependency cycles, SOLID compliance.
AI Cost Optimization
Set up intelligent model routing. Route simple tasks to cheap models, complex tasks to powerful ones. Measurable savings.
Agent Ecosystem Setup
Configure the full pipeline: quality gates, cost tracking, content filtering. Production-ready in days.
Live Scan: archlint on itself
archlint-rs scans its own Go + Rust codebase. Full results