HN Showcase
Back to browse
GitHub Repository

Stop babysitting your agent. Architecture rules it can't ignore, checked on every change before it moves on.

31 starsTypeScript

Yggdrasil – Scoped architectural rule enforcement for AI-generated code

by chrisdudek·Apr 14, 2026·1 point·0 comments
Visit ProjectView on HN

AI Analysis

●●●BangerBig BrainSolve My Problem

Finally enforces .cursorrules that AI agents normally ignore — runs before PR review.

Strengths
  • •Reviewer LLM checks agent output against rules while coding, not after PR
  • •Scoped rules show agents only 3-5 relevant constraints per file
  • •Markdown rules are human-readable, not a custom DSL
Weaknesses
  • •Requires Node.js 22+ which may block some enterprise environments
  • •Reviewer LLM calls add latency to the agent workflow
Target Audience

Teams using AI coding agents like Cursor or Claude

Similar To

Cursor rules · CLAUDE.md · Sourcegraph Cody

Similar Projects

Security●●Solid

Access-aware text-to-SQL – stop LLM agents overfetching data

Deterministic guard prevents overfetching when 121 adversarial tests pass.

Big BrainNiche Gem
dimitarst
1113d ago
Developer Tools●●Solid

Personal AI gateway for OpenClaw – tokenomics

OpenAI-compatible proxy with PII masking and token budgets—but LiteLLM, Helicone already do this.

Solve My ProblemBig Brain
crawdog
204mo ago
AI/ML●●Solid

Open Bias – proxy that enforces agent behavior at runtime

Local proxy enforcing markdown rules on LLM output before it hits production.

Solve My ProblemShip It
algomaniac
2142mo ago