Integrations

Connect Falcon with your IDE, CI/CD pipeline, and third-party tools via MCP, HTTP API, Rust SDK, and LSP.

1. MCP Server

Falcon provides an MCP (Model Context Protocol) server that exposes Falcon's analysis capabilities to AI assistants like Cursor. Run it with falcon-mcp or falcon x mcp.

What is MCP?

MCP (Model Context Protocol) is a standard that lets AI tools discover and invoke external tools. Falcon's MCP server exposes static analysis, scoring, and fix capabilities so your AI assistant can analyze code, explain rules, and suggest fixes directly in context.

Configure in Cursor

Add Falcon to your Cursor MCP settings (~/.cursor/mcp.json or project-level config):

{
  "mcpServers": {
    "falcon": {
      "command": "falcon-mcp"
    }
  }
}

Available Tools

Falcon exposes 7 MCP tools:

• falcon_analyze — Full project analysis
• falcon_ai_score — AI-powered quality score
• falcon_check_file — Check a single file
• falcon_explain_rule — Explain a rule's purpose and fix
• falcon_fix — Apply automated fixes
• falcon_conventions — List project conventions
• falcon_provenance — Show AI provenance for code

Example: tools/list

// JSON-RPC request
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/list"
}

// Response (simplified)
{
  "jsonrpc": "2.0",
  "id": 1,
  "result": {
    "tools": [
      { "name": "falcon_analyze", "description": "Analyze project..." },
      { "name": "falcon_check_file", "description": "Check single file..." }
      // ... 5 more
    ]
  }
}

Example: tools/call with falcon_check_file

// Request
{
  "jsonrpc": "2.0",
  "id": 2,
  "method": "tools/call",
  "params": {
    "name": "falcon_check_file",
    "arguments": {
      "path": "lib/main.dart",
      "source": "void main() { print('hi'); }"
    }
  }
}

// Response
{
  "jsonrpc": "2.0",
  "id": 2,
  "result": {
    "content": [
      { "type": "text", "text": "{\"issues\":[],\"score\":100}" }
    ]
  }
}

2. HTTP API

Run Falcon as an HTTP server with falcon x api --port 8090 to integrate with webhooks, custom dashboards, or other services.

Endpoints

Example: POST /analyze

curl -X POST http://localhost:8090/analyze \
  -H "Content-Type: application/json" \
  -d '{"source": "void main() { print(\"hello\"); }"}'

Example: POST /score

curl -X POST http://localhost:8090/score \
  -H "Content-Type: application/json" \
  -d '{"project_path": "/path/to/flutter/project"}'

Example: GET /health

curl http://localhost:8090/health

// Response
{
  "status": "ok",
  "version": "0.1.0"
}

3. Rust SDK

Use Falcon programmatically from Rust via the falcon::sdk crate.

Add as dependency

// Cargo.toml
[dependencies]
falcon = { git = "https://github.com/viveky259259/falcon" }

Analyze project

use falcon::sdk::{analyze_project, AnalyzeOptions};

let result = analyze_project("/path/to/project", AnalyzeOptions::default())?;
for issue in result.issues {
    println!("{:?}", issue);
}

Analyze single file

use falcon::sdk::analyze_source;

let source = r#"void main() { print('hi'); }"#;
let result = analyze_source(source)?;

Score project

use falcon::sdk::score_project;

let score = score_project("/path/to/project")?;
println!("Score: {}/100", score.total);

Analyze to JSON

use falcon::sdk::analyze_to_json;

let json = analyze_to_json("/path/to/project")?;
println!("{}", json);

4. LSP Server

Falcon provides an LSP (Language Server Protocol) server (falcon-lsp) for real-time diagnostics and code actions in your editor.

VS Code extension setup

Install the Falcon VS Code extension from the marketplace, or build from source. The extension automatically discovers falcon-lsp on your PATH.

Features

• Diagnostics — Inline squiggles for rule violations as you type
• Code actions — Quick fixes for auto-fixable rules
• AI provenance hints — Inlay hints showing which rules flagged AI-generated code

Configuration

Add to your settings.json:

{
  "falcon.enabled": true,
  "falcon.lspPath": "falcon-lsp",
  "falcon.rules": ["prefer_const", "avoid_print"]
}

5. Partner Integrations