Tivra DebugMind for VS Code

Your AI debugging copilot for production services.

Automatically fetch production logs, get AI-powered root cause analysis, and receive actionable fixes — all without leaving VS Code.

Philosophy

"From production error to fix in 5 minutes, not 4 hours"

Stop context-switching between log dashboards, consoles, and code. Tivra DebugMind brings production debugging directly into your IDE with intelligent log analysis and AI-powered root cause analysis.

What's New in v0.1.0 🎉

Initial Release - AI Debugging Copilot for Production Services

🎯 EC2 & ECS Support: Monitor virtual machines and containerized services
🔍 Intelligent Log Analysis: Automatically fetch and analyze production logs
✅ Smart "No Errors" Detection: Know when everything is working perfectly
🤖 AI-Powered RCA: Claude AI analyzes logs and provides actionable fixes
✨ Agentic Code Fixes: Apply AI-generated fixes directly in your IDE with one click
⚡ Multi-Service Scanning: Analyze errors across all services in one click
💬 Conversational Interface: Chat with AI about production errors
🔄 Auto-Discovery: Services automatically detected on startup

Features

🔗 Production Debugging: Connect to your cloud services and fetch real production logs
🔍 Intelligent Log Analysis: Automatically scan logs for errors, exceptions, and failures
🤖 AI Root Cause Analysis: Get detailed analysis of error patterns and suggested fixes
✨ Agentic Code Fixes: One-click application of AI-generated fixes directly in your workspace
- Review suggested code changes with inline diffs
- Apply fixes to your codebase instantly
- Full undo support
- No manual copy-pasting required
✅ Zero False Positives: Get positive feedback when services are healthy
⚡ One-Click Analysis: Scan all services for errors with a single click
💬 Conversational Debugging: Ask questions about errors and get instant insights
🔄 Automatic PR Creation: Generate pull requests with fixes ready for review
✅ Zero Context Switching: Everything in VS Code - no log dashboards, no cloud consoles

Why Tivra DebugMind?

The Problem:

Production errors require jumping between cloud consoles, log dashboards, and code
Searching through thousands of log lines manually
Copying error messages to Google/Stack Overflow
Multiple deploy cycles to test fixes
Hours of context switching and debugging toil

The Solution:

┌─────────────────┐
│  Write Code     │
└─────────────────┘
        ↓
┌─────────────────┐
│  Deploy         │
└─────────────────┘
        ↓
┌─────────────────┐
│  Error Occurs   │ ← Production logs
└─────────────────┘
        ↓
┌─────────────────┐
│  DebugMind      │ → Fetch logs → Analyze → Suggest fix
└─────────────────┘
        ↓
┌─────────────────┐
│  Apply Fix      │ → Back to code
└─────────────────┘

Tivra DebugMind reduces debugging toil by bringing production context directly into your IDE.

Quick Start

1. Install Extension

Download the .vsix file from GitHub Releases and install:

Open VS Code
Press Cmd+Shift+P (Mac) or Ctrl+Shift+P (Windows/Linux)
Type: Extensions: Install from VSIX...
Select the downloaded .vsix file
Reload VS Code

2. Open Tivra DebugMind Copilot

Click the DebugMind button in the VS Code status bar (bottom left)
- Or open Command Palette (Cmd+Shift+P) and run: Tivra DebugMind
The AI copilot chat panel will open

3. Connect to Your Services

In the chat panel:

Click the suggested prompt "Connect me to AWS"
Enter your credentials when prompted:
- Access Key ID
- Secret Access Key
- Region (e.g., us-east-1, ap-south-1)
The extension will automatically discover your EC2 and ECS services

4. Analyze Production Logs

Click "Analyze errors in my services" from the suggested prompts
The AI will:
- Automatically fetch production logs from ALL your services
- Detect errors, exceptions, and failures from the last hour
- Group errors by service and severity
- Perform root cause analysis
- Suggest actionable code fixes

If no errors are found:

✅ No Errors Found!
All your services are running smoothly with no errors in the last hour.
Everything looks perfect! 🎉

If errors are detected:

⚠️ Errors Found
Found 5 error(s) across 2 service(s)

### payment-api (ecs)
**3 error(s)**
1. NullPointerException in payment processing
2. Database connection timeout

🔍 Root Cause Analysis
The NullPointerException occurs when processing payments without
a payment method specified. Add null validation before accessing
payment.method property.

Suggested Fix:
[Code changes with explanations]

5. Chat About Production Issues

Ask the copilot anything about production errors:

"Why is my payment-api timing out?"
"Show me errors from my EC2 instances"
"What's causing the NullPointerException?"
"How do I fix the database connection issue?"

Complete Workflow Example

1. Deploy new payment processing feature to production

2. Error occurs in production ECS service "payment-api"
   └─ NullPointerException (47 occurrences in logs)

3. Open Tivra DebugMind copilot in VS Code
   └─ Click "Analyze errors in my services"
   └─ AI fetches production logs automatically

4. AI Root Cause Analysis:
   ├─ Error: "NullPointerException in PaymentProcessor.java:142"
   ├─ Root Cause: "Missing null check for optional payment method"
   ├─ Suggested Fix: Add null validation and default handling
   └─ Shows exact code changes with inline diff

5. Apply the AI-generated fix:
   ├─ Click "Apply Fix" button in the chat
   ├─ Code automatically updated in your workspace
   ├─ Review the changes with VS Code's diff view
   ├─ Optionally click "Create PR" to generate pull request
   └─ Deploy → Total debugging time: 5 minutes (vs 2-4 hours manual)

How It Works

Traditional Debugging (2-4 hours)

1. Get alert about production error
2. Open cloud console / log dashboard
3. Search through thousands of log lines
4. Copy error messages to Google/Stack Overflow
5. Guess at the fix
6. Try fix [#1](https://github.com/shrey2525/tivra-copilot-vscode/issues/1) → Deploy → Wait → Still failing
7. Try fix [#2](https://github.com/shrey2525/tivra-copilot-vscode/issues/2) → Deploy → Wait → Still failing
8. Try fix [#3](https://github.com/shrey2525/tivra-copilot-vscode/issues/3) → Deploy → Wait → Success!
   Total: 2-4 hours, 15+ context switches, 4+ deployments

With Tivra DebugMind (5 minutes)

1. Open DebugMind copilot in VS Code
2. Click "Analyze errors in my services"
3. AI fetches logs, performs RCA, suggests fix
4. Apply fix → Deploy → Success!
   Total: 5 minutes, 0 context switches, 1 deployment

Commands

Tivra DebugMind - Open the AI copilot chat panel
Tivra: Start Debugging - Quick start debugging with service selection
Tivra: Connect to AWS - Connect your cloud account (also available in chat)

In the Chat Panel:

Click "Connect me to AWS" - Connect your account
Click "Analyze errors in my services" - Scan all services for errors
Type any question about your production errors
Use suggested prompts for common debugging tasks

Configuration

The extension uses the production API by default (https://copilot.tivra.ai).

For local development, you can override the API URL in your VS Code settings:

{
  "tivra.apiUrl": "http://localhost:3001"
}

Then run the backend server locally:

cd tivra-copilot/server
npm install
node index.js

Requirements

VS Code 1.85.0+
Cloud account with log access (currently supports AWS)
API credentials for log fetching

Privacy & Security

Credentials are stored securely in VS Code's Secret Storage
Logs are analyzed in real-time, not stored permanently
All fixes require your explicit review and approval
Full undo support via VS Code's built-in undo

Supported Services (v1)

✅ Amazon EC2 (Virtual machines)
✅ Amazon ECS (Containerized services - Fargate & EC2)

Coming in future versions:

🔜 AWS Lambda
🔜 Amazon RDS
🔜 Kubernetes
🔜 Google Cloud Platform
🔜 Azure

Roadmap

v1.0 (Current)

✅ EC2 and ECS service discovery
✅ Log analysis and error detection
✅ AI-powered root cause analysis
✅ Conversational debugging interface

v2.0 (Planned)

Lambda and RDS support
Kubernetes integration
Multi-cloud support (GCP, Azure)
Performance metrics analysis
Distributed tracing

v3.0 (Future)

Automatic PR generation with fixes
Deployment rollback suggestions
Cost optimization recommendations
Custom alerting rules

Use Cases

1. Debugging Production Errors

Your code works locally but breaks in production
Tivra fetches production logs and shows you exactly what's failing

2. Post-Deployment Validation

Just deployed new code
Run a quick health check to ensure no new errors

3. On-Call Debugging

Got paged about an error
Quickly analyze logs and get AI-suggested fixes

4. Reducing Debugging Toil

Stop jumping between cloud consoles and log dashboards
Get all debugging context in one place

Feedback & Support

🐛 Report issues: GitHub Issues
💬 Feature requests: GitHub Discussions
📧 Email: info@tivra.ai

License

MIT License - see LICENSE for details

Made with ❤️ by the Tivra team

Reducing debugging toil, one log at a time.

Tivra DebugMind

shrey chaturvedi