Rse CLI: `ai` command

Note: This command is currently in the development and may have some limitations. This README.md will be updated over time. Feedback is welcome!

Discord — Docs — Sponsors

@reliverse/ai is a multi-modal AI agent. This integration allows for AI-powered chat and agents. Easily invoke specialized AI agents using simple chat commands and watch them work magic on your code.

Features

• Chat-based AI interactions
  Invoke agentic tools in a single chat message using @agent-name with automatically detected parameters like paths to files or directories.

• 🧩 Agent-based architecture
  Each agent targets a specific problem domain (e.g., AI linting).

• ⚡ Context-aware processing
  Free-form text around agent calls is combined into contextual input for more accurate AI responses.

• 🤖 @relinter
  Offers AI-driven linting checks that go beyond traditional rule-based linters.

• 🏗️ Seamless CLI integration
  Ties directly into @reliverse/rse for easy project setup and usage.

• 📝 Generate Code with AI
  Use rse ai code <prompt> to generate code based on your prompts.

• 🤖 MCP (Model Context Protocol) "All the useful things in the world, like Slack, GitHub, or even your local filesystem, provide their own unique API to access them. LLM's can call these API's using tools - but these tools require you to write code to connect the LLM to the API. And if the LLM is inside a desktop application, like Cursor or Claude Desktop, you can't manually add new tools yourself. Wouldn't it be nice if you could quickly get access to a set of pre-made tools? And even better, could they work with my existing desktop apps? The Model Context Protocol lets you build toolsets which can be plugged into existing applications." © Matt Pocock

Chat Usage

Chat commands follow this pattern:

@agentName <some optional free-form text> @-some-parameter

For example:

@relinter please lint code quality in @-src/components

All text before, after, or between these tokens (@agentName, @-params) is added as context for the agent to use.

Agents

1. Reimgen Agent

Reimgen is an AI-powered image generation tool that can help you create images for your project.

# Simple usage
rse ai gen "robot playing guitar"

# Change model
rse ai gen "watercolor forest" -m stable-diffusion

# Specify provider for upload
rse ai gen "space sunset" -p uploadcare

# Experimental chat invocation
@reimgen gen an image of a cat

2. Relinter Agent

Relinter is an AI-powered alternative to classic linting tools like ESLint—built to spot a wide range of code issues and deliver suggestions in a convenient JSON report.

Usage Example

@relinter please lint code quality in @-src/components

How It Works

• AI Processing: The specified files are analyzed by rse AI (e.g., GPT-4o-mini).
• Output: Results are stored in relinter.json, featuring:
  • File paths
  • Precise issue locations (start/end lines)
  • AI-suggested improvements

What It Detects

• Code quality issues: type safety, unused variables, risky equality checks, and more
• Architectural concerns: circular dependencies, modular structure problems
• Context-specific improvements, tailored to your codebase

ESLint-Like Issues

// relinter-test-a.ts
function add(a: any, b: any) {
  return a + b;
}

const result = add(5, "10");
console.log("The result is:", result);

let unusedVar = 123;

if (result == 510) {
  console.log("Result equals 510");
}

Possible lint findings:

• Type coercion: mixing number and string
• Unused variables: unusedVar
• Loose equality: == instead of ===

Circular Dependencies Issue

// relinter-test-b.ts
import { functionC } from "./relinter-test-c.js";

export function functionB() {
  console.log("Function B called");
  functionC();
}

// relinter-test-c.ts
import { functionB } from "./relinter-test-b.js";

export function functionC() {
  console.log("Function C called");
  functionB();
}

Detected issue:

• Circular import between relinter-test-b.ts and relinter-test-c.ts

TODO

• Implement basic rse AI Relinter Agent functionality
• Provide a VSCode-like extension to visualize Relinter suggestions inline
• Customize severity thresholds (error, warning, info) in relinter.json
• Enable quick-fix suggestions for auto-applying changes
• Integrate with CI pipelines and pre-commit hooks automatically
• Implement MCP which allows AIs to interact with rse CLI
• Allow developers connect their MCP to extend rse CLI
• Connect rse CLI to Glama MCP servers
• ai code should follow .rse/rules/*.{md,mdc}

Show Your Love

If you find this helpful, please consider supporting us or joining our Discord community.