RuleForge | Sorted. Tools

RULEFORGE / AGENTS.md ◀ → CLAUDE.md → .cursorrules → SKILL.md

Quick tip. Fill in as much as you can — the more context, the more precise and immediately usable the generated config. Most fields are optional; only format, project name, description, and tech stack are required.

Waiting for input...

READINESS REPORT

OUTPUT FORMATS

Target Format(s) Select one or more — each generates a separate prompt

Recommended Model

PROJECT IDENTITY

Project Name Required

Monorepo?

One-Line Description Required — loaded into every AI session

💬 Keep it concise — this is loaded into every session. Aim for under 150 characters.

💬 Avoid vague adjectives — be specific about what this project does.

Project Purpose Optional — helps AI make better judgment calls

TECH STACK

Language Required — select all that apply

Framework Required — select all that apply

Database Optional

Package Manager

Additional Tools Optional

Deployment Optional

COMMANDS & WORKFLOW

💬 AI will guess commands without these — errors likely. Enter at least a dev and build command.

Dev Server

Build

Test

Lint / Format

Deploy

💬 Avoid sudo in AI agent commands — use project-scoped tools instead.

Type Check

Other Commands Optional — key=value format

Verification Workflow

CODE STYLE & CONVENTIONS

💡 Tip: Less is more. Only specify styles your linter doesn't catch.

Module System

Export Style

Indentation

Quotes

Semicolons

Commit Format

Naming Convention Optional

Linter / Formatter Optional

Testing Framework Optional

Test Pattern Optional

Additional Style Notes Optional

ARCHITECTURE & BOUNDARIES

Key Directories Optional

Protected Files Files AI must NEVER modify

💬 Consider protecting .env and config files — they're easy to accidentally modify.

Key Architectural Rules Optional

Important Gotchas Optional — the more specific, the fewer AI mistakes

AI Permissions

Forbidden Actions Optional

FORMAT-SPECIFIC OPTIONS