@hivehub/rulebook
Tool-agnostic AI development framework. Standardize projects across Claude Code, Cursor, Gemini, Codex, Windsurf, Copilot with automated templates, quality gates, persistent memory, and framework detection for 28 languages, 17 frameworks, 13 MCP modules, and 20 services.
Why Rulebook?
Large Language Models (LLMs) for software development need clear directives to generate high-quality code consistently. Without standardized guidelines, AI-generated code can be inconsistent, error-prone, and difficult to maintain.
Rulebook solves this by providing:
- 📋 Comprehensive Rule Sets: Language-specific guidelines (TypeScript, Rust, Python, etc.), framework conventions (NestJS, Django, React), testing standards, linting rules, spell-checking, CI/CD pipelines, Git hooks, and version control best practices
- 🤖 Ralph Autonomous Loop: Multi-iteration AI agent task solving with quality gates (type-check, lint, tests, coverage), iteration tracking, learnings extraction, and graceful pause/resume
- 🧠 Persistent Memory: Context that survives across AI sessions with hybrid BM25+HNSW search, native SQLite with WASM fallback
- 🎯 Normalized Deliverables: Ensures consistent code quality across all AI-generated outputs
- 🛡️ Error Reduction: Catches issues early through automated quality gates and pre-commit/pre-push hooks
- ⚙️ Process Automation: Automates repetitive tasks like formatting, testing, and deployment
- 🔌 MCP Integration: 40 MCP tools for task management, skills, memory, decisions, knowledge, and learnings
- 📊 Task Management: Built-in task management system (OpenSpec-compatible format) for structured development
By giving LLMs a clear "rulebook" to follow, you ensure that every piece of generated code meets your project's standards—automatically.
Quick Start
# New project — auto-detects languages, tools, and complexity
npx @hivehub/rulebook@latest init
# Target specific AI tools
npx @hivehub/rulebook@latest init --tools claude-code,cursor,gemini
# Assess project complexity before configuring
npx @hivehub/rulebook@latest assess
# Minimal setup (essentials only)
npx @hivehub/rulebook@latest init --minimal
# Lean mode — AGENTS.md as <3KB index (fast AI loading)
npx @hivehub/rulebook@latest init --lean
# Update existing project (re-projects rules to all detected tools)
npx @hivehub/rulebook@latest update
# Manage canonical rules
npx @hivehub/rulebook@latest rules list # List rules by tier
npx @hivehub/rulebook@latest rules add no-shortcuts # Install from library
npx @hivehub/rulebook@latest rules project # Project to all tools
What's New
See the full CHANGELOG for details.
v5.2.0 — Phase-Based Task Organization
Tasks are now organized by phases with enforced naming: phase<N>_<description> (e.g., phase0_setup, phase3a_fix-auth). Inspired by real-world project management where priority and execution order must be visible at a glance in the file explorer.
Phase Naming Convention
- Task IDs MUST start with
phase<N>[subletter]_prefix — rejected at creation otherwise - Phases define execution order and priority:
phase0→phase1→phase2a→phase2b→ ... - Tasks must fit a single implementation cycle — large tasks are split into multiple phase-prefixed tasks
Auto-Generated Tasks Index
tasks/README.mdauto-generated on every task create, archive, delete, and status update- Tasks grouped by phase with status icons, progress counters (
3/5), and descriptions