Elemental Cursor Rules - Dynamic Rule Fetching

This skill enforces a strict workflow in Cursor IDE that requires fetching and applying project-specific rules before every AI response or code modification.

What This Skill Does

This skill establishes a mandatory rule-checking protocol that ensures all AI responses and code changes comply with project-specific guidelines. It requires the AI assistant to fetch rules dynamically at multiple checkpoints throughout development.

Instructions for AI Assistant

CRITICAL: Mandatory Rule Fetching Protocol

**You MUST call `fetch_rules` before EVERY action. No exceptions.**

When to Fetch Rules

1. **At conversation start** - Always fetch rules when a new conversation begins

2. **Before EVERY user request** - Fetch rules before processing any user input

3. **Before ANY code changes** - Never modify code without first fetching applicable rules

4. **When topic changes** - Fetch rules again if the conversation shifts to a different feature or area

5. **During active conversations when**:

- User mentions a new feature or component

- File type changes (e.g., switching from .py to .tsx)

- Task type changes (e.g., from debugging to adding features)

Rule Application Process

1. **Fetch rules** using the `fetch_rules` command/tool

2. **Check ALL matching rules**:

- General guidelines (ALWAYS apply these)

- Feature-specific rules (if working on a particular feature)

- File type rules (based on file extensions like .py, .mdc, .tsx, etc.)

3. **Apply rules** - Let rules override your general knowledge

4. **Proceed** - Only after applying all relevant rules, proceed with the response or code change

Rule Priority

**Project rules ALWAYS override your base knowledge.** If a project rule conflicts with general best practices, follow the project rule.

Compliance Requirements

Example Usage

Scenario 1: Starting a New Conversation

```

User: "I need to add a new API endpoint"

Assistant: [Calls fetch_rules] → [Reviews general, API, and backend rules] → Responds with compliant approach

```

Scenario 2: Mid-Conversation Topic Shift

```

User: "Actually, let's work on the frontend instead"

Assistant: [Calls fetch_rules again] → [Reviews general, frontend, and React rules] → Responds with compliant approach

```

Scenario 3: File Type Change

```

User: "Now let's update the Python data processing script"

Assistant: [Calls fetch_rules again] → [Reviews general and .py file rules] → Responds with compliant approach

```

Important Notes

Constraints