AI SummaryLár Windsurf Rules is a cursor ruleset that guides developers in building graph-based AI agents using the Lár framework with strict typing, explicit node linking, and auditable patterns. It benefits developers building complex, self-correcting agentic systems who need clear structural guidelines.
Install
Copy this and paste it into Claude Code, Cursor, or any AI assistant:
I want to add the "lar — Windsurf Rules" prompt rules to my project. Repository: https://github.com/snath-ai/lar Please read the repo to find the rules/prompt file, then: 1. Download it to the correct location (.cursorrules, .windsurfrules, .github/prompts/, or project root — based on the file type) 2. If there's an existing rules file, merge the new rules in rather than overwriting 3. Confirm what was added
Description
Lár the Pytorch for Agents is the open-source "glass box" engine for building, debugging, and running auditable, self-correcting AI agents.
Lár Framework - Cursor Rules
You are coding with Lár, a graph-based agent framework that treats "Code as the Graph".
Core Principles
• Strict Typing: Every Node and Function MUST have Pydantic/Python type hints. • Explicit Linking: Connect nodes using .next_node = target or RouterNode(path_map={...}). • No "Magic": Do not assume global state. Use state.get() and state.set().
1. Defining a Node
`python from lar import LLMNode architect = LLMNode( model_name="gemini/gemini-1.5-pro", prompt_template="Analyze {input}...", output_key="plan" ) `
2. Defining a Tool
`python from lar import ToolNode def my_tool(state): return "result" tool = ToolNode( tool_function=my_tool, input_keys=["__state__"], output_key="result", next_node=None ) `
Discussion
Health Signals
My Fox Den
Community Rating
Sign in to rate this booster
Works With
Any AI assistant that accepts custom rules or system prompts