Argus Server

Release License: MIT

Model Context Protocol (MCP) server for Argus E2E Testing Agent. This allows AI coding assistants to interact with Argus testing capabilities directly from your IDE.

Supported AI IDEs

• Claude Code - Anthropic's CLI for Claude
• Cursor - AI-first code editor
• Windsurf - Codeium's AI IDE
• VS Code - With MCP extension
• Any MCP-compatible client

Available Tools (74 Total)

Authentication (4 tools)

Tool	Description
argus_auth	Start OAuth2 device flow authentication
argus_auth_complete	Complete OAuth2 device flow after sign-in
argus_auth_status	Check current authentication status
argus_auth_logout	Sign out and clear stored credentials

Core Browser Testing (4 tools)

Tool	Description
argus_health	Check Argus API status
argus_discover	Discover interactive elements on a page
argus_act	Execute browser actions (click, type, navigate)
argus_test	Run multi-step E2E tests with screenshots

Data Extraction & AI Agent (3 tools)

Tool	Description
argus_extract	Extract structured data from pages
argus_agent	Autonomous task completion on websites
argus_generate_test	Generate test steps from natural language

Quality Intelligence (3 tools)

Tool	Description
argus_quality_score	Get overall project quality score (0-100)
argus_quality_stats	Get detailed quality statistics
argus_risk_scores	Calculate/retrieve risk scores for components

Production Events & Analysis (3 tools)

Tool	Description
argus_events	List production errors/events with filtering
argus_event_triage	AI-powered triage recommendations for events
argus_test_from_event	Generate E2E test from production error

Test Management (3 tools)

Tool	Description
argus_tests	List generated tests by status
argus_test_review	Review (approve/reject/modify) generated tests
argus_batch_generate	Generate tests for multiple events at once

Self-Healing (4 tools)

Tool	Description
argus_healing_config	View/update self-healing configuration
argus_healing_patterns	View learned self-healing patterns with confidence scores
argus_healing_stats	Get self-healing statistics (success rates, patterns)
argus_healing_review	Approve/reject healing suggestions

Synchronization (4 tools)

Tool	Description
argus_sync_push	Push local test changes to cloud
argus_sync_pull	Pull test updates from cloud
argus_sync_status	Check sync status and conflicts
argus_sync_resolve	Resolve synchronization conflicts

Test Export & Recording (4 tools)

Tool	Description
argus_export	Export test to code (Python/TS/Java/C#/Ruby/Go)
argus_export_languages	List supported export languages/frameworks
argus_recording_to_test	Convert browser recording (rrweb) to test
argus_recording_snippet	Generate JavaScript snippet for recording sessions

Collaboration (2 tools)

Tool	Description
argus_presence	Get/update user presence in workspace
argus_comments	Get/add/reply to test comments

Project Management (4 tools)

Tool	Description
argus_projects	List projects user has access to
argus_what_to_test	Get AI recommendations on what to test next
argus_coverage_gaps	Identify areas with production errors but no tests
argus_dashboard	Get comprehensive project dashboard/overview

Infrastructure & Monitoring (5 tools)

Tool	Description
argus_infra_overview	Get infrastructure costs and Selenium Grid status
argus_infra_recommendations	Get AI recommendations for cost optimization
argus_infra_apply	Apply infrastructure optimization recommendation
argus_browser_pool	Get Selenium Grid browser pool real-time status
argus_llm_usage	Get LLM/AI model usage and costs

AI Assistant (1 tool)

Tool	Description
argus_ask	Ask AI questions about your tests/errors/strategy

CI/CD Integration (4 tools) 🆕

Tool	Description
argus_cicd_test_impact	AI-powered test impact analysis for PRs - 10x faster CI
argus_cicd_deployment_risk	Risk assessment before deployments
argus_cicd_builds	List build history
argus_cicd_pipelines	View CI/CD pipeline status

Discovery & Intelligent Crawling (4 tools) 🆕

Tool	Description
argus_discovery_start	Start autonomous app crawling session
argus_discovery_flows	Get AI-discovered user flows
argus_discovery_generate	Generate tests from discovered flows
argus_discovery_compare	Compare discovery sessions

Time Travel Debugging (5 tools) 🆕

Tool	Description
argus_time_travel_checkpoints	List state checkpoints for test runs
argus_time_travel_history	View state change history for a thread
argus_time_travel_replay	Replay execution from a checkpoint
argus_time_travel_fork	Fork execution for A/B testing
argus_time_travel_compare	Compare divergent execution paths

Visual AI Testing (5 tools) 🆕

Tool	Description
argus_visual_capture	Capture screenshot with metadata
argus_visual_compare	AI-powered visual diff between screenshots
argus_visual_baseline	Set a visual baseline
argus_visual_baselines	List visual baselines
argus_visual_analyze	Analyze visual changes with WCAG accessibility

Scheduling & Automation (4 tools) 🆕

Tool	Description
argus_schedule_create	Create a test schedule (cron-style)
argus_schedule_list	List schedules
argus_schedule_run	Manually trigger a schedule
argus_schedule_history	View run history for a schedule

Correlation & Analytics (4 tools) 🆕

Tool	Description
argus_correlations_timeline	View unified SDLC event timeline
argus_correlations_root_cause	AI-powered root cause analysis
argus_correlations_insights	Get AI-generated insights
argus_correlations_query	Natural language queries for correlations

API Testing (3 tools) 🆕

Tool	Description
argus_api_discover	Discover API endpoints from OpenAPI/Swagger
argus_api_generate	Generate API test cases with AI
argus_api_run	Execute API tests

Installation

For Claude Code / Claude Desktop

Add to your MCP settings (~/.claude/claude_desktop_config.json or Claude Code settings):

{
  "mcpServers": {
    "argus": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://argus-mcp.samuelvinay-kumar.workers.dev/sse"]
    }
  }
}

For Cursor

Add to Cursor's MCP settings:

{
  "mcpServers": {
    "argus": {
      "url": "https://argus-mcp.samuelvinay-kumar.workers.dev/sse"
    }
  }
}

Usage Examples

Discover Page Elements

Use argus_discover to analyze https://example.com and find all interactive elements

Run E2E Test

Use argus_test to test login on https://uitestingplayground.com/sampleapp with these steps:
1. Type "TestUser" in the username field
2. Type "pwd" in the password field
3. Click the Log In button
4. Verify the welcome message appears

AI-Powered Test Impact Analysis

Use argus_cicd_test_impact to analyze which tests need to run for my PR changes

Autonomous App Discovery

Use argus_discovery_start to crawl https://myapp.com and discover all user flows

Visual Regression Testing

Use argus_visual_compare to compare the current homepage against the baseline

Time Travel Debugging

Use argus_time_travel_checkpoints to list all state snapshots for the failed test run

Root Cause Analysis

Use argus_correlations_root_cause to analyze why the checkout test started failing

API Testing

Use argus_api_discover to find all endpoints from our OpenAPI spec, then generate tests

Development

Prerequisites

• Node.js 18+
• Wrangler CLI

Setup

cd argus-mcp-server
npm install

Local Development

npm run dev

Deploy

npm run deploy

Architecture

┌─────────────────────────────────────────────────────────────┐
│                    AI IDE (Claude/Cursor)                    │
└─────────────────────────┬───────────────────────────────────┘
                          │ MCP Protocol (SSE)
                          ▼
┌─────────────────────────────────────────────────────────────┐
│                   Argus MCP Server                           │
│              (Cloudflare Workers + Durable Objects)          │
│                                                              │
│  ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐       │
│  │  74 MCP  │ │   OAuth  │ │    R2    │ │   KV     │       │
│  │  Tools   │ │  Auth    │ │ Storage  │ │  State   │       │
│  └────┬─────┘ └────┬─────┘ └────┬─────┘ └────┬─────┘       │
└───────┼────────────┼────────────┼────────────┼──────────────┘
        │            │            │            │
        └────────────┴────────────┴────────────┘
                          │ REST API
                          ▼
┌─────────────────────────────────────────────────────────────┐
│                     Argus Backend                            │
│                                                              │
│  ┌────────────┐  ┌────────────┐  ┌────────────┐            │
│  │ Browser    │  │ AI Brain   │  │ Knowledge  │            │
│  │ Pool (K8s) │  │ (Railway)  │  │ Graph      │            │
│  └────────────┘  └────────────┘  └────────────┘            │
└─────────────────────────────────────────────────────────────┘

Features

Self-Healing Selectors

AI-powered selector recovery when elements change

Smart Waits

Intelligent element detection without explicit waits

Visual AI Testing

Screenshot comparison with WCAG accessibility analysis

Time Travel Debugging

Replay test executions from any checkpoint

CI/CD Integration

AI-powered test impact analysis for 10x faster pipelines

Multi-Browser Support

Chrome, Firefox, Safari, Edge via Selenium Grid

License

MIT