Qc Backtest Runner

QuantConnect Cloud Backtest Runner

Description

This skill automates the complete workflow for running backtests on QuantConnect cloud:

1. Optional git commit and push to GitHub (git@github.com:vinamra2013/QuantConnectAlgoTradingStratigies.git)
2. Push strategy code to QuantConnect cloud via LEAN CLI
3. Execute backtest on QuantConnect servers
4. Wait for completion and retrieve results
5. Parse and display key performance metrics

When to Use

Auto-activate when user requests:

• "Run backtest on QuantConnect"
• "Test strategy on QC cloud"
• "Upload and backtest on QuantConnect"
• "Run cloud backtest"
• Keywords: "quantconnect", "qc cloud", "lean backtest", "cloud backtest"

Manual invocation:

• User explicitly asks to run QuantConnect backtest
• After strategy modifications when validation is needed
• When comparing strategy performance

Prerequisites

1. LEAN CLI installed:

   pip install lean
   lean login
   

2. Environment variables in .env:

   QC_PROJECT_NAME="Your Project Name"
   

3. GitHub repo configured:

   • Repo: git@github.com:vinamra2013/QuantConnectAlgoTradingStratigies.git
   • Strategy files in strategies/ directory

Parameters

Required

• None (uses defaults from .env)

Optional

• --project: QuantConnect project name (default: from QC_PROJECT_NAME env var)
• --commit: Commit and push to GitHub before backtest
• --commit-message: Custom git commit message
• --open: Open results in browser after starting backtest
• --wait: Wait for backtest to complete (polls every 10s)
• --save: Download and save results locally to results/qc/
• --timeout: Wait timeout in seconds (default: 300)
• --no-push: Skip pushing to QC cloud (use existing code)

How to Use This Skill

The skill includes a bundled script scripts/qc_cloud_backtest.py that automates the entire backtest workflow.

Quick Backtest (Start Only)

Execute the bundled script to start a backtest and open results in browser:

venv/bin/python  .claude/skills/qc-backtest-runner/scripts/qc_cloud_backtest.py --open

Full Workflow (Wait for Results)

Execute with wait and save flags to retrieve complete results:

venv/bin/python  .claude/skills/qc-backtest-runner/scripts/qc_cloud_backtest.py --wait --save

With Git Commit

Commit and push changes before running backtest:

venv/bin/python  .claude/skills/qc-backtest-runner/scripts/qc_cloud_backtest.py --commit --wait --save

Custom Project

Specify a different QuantConnect project:

venv/bin/python .claude/skills/qc-backtest-runner/scripts/qc_cloud_backtest.py --project "My Custom Strategy" --open

Output

Immediate Feedback

• ✓ LEAN CLI version
• ✓ Project name confirmation
• ✓ Git push status (if --commit)
• ✓ Cloud push status
• ✓ Backtest started confirmation

When Using --wait --save

============================================================
BACKTEST RESULTS
============================================================
Total Return        : 392.23%
Sharpe Ratio        : 0.829
Total Orders        : 2
Win Rate            : 0%
Loss Rate           : 0%
Average Win         : 0%
Average Loss        : 0%
Max Drawdown        : 59.000%
Sortino Ratio       : 0.856
Net Profit          : $3922.28
Annual Return       : 37.501%
============================================================

Result Files

• JSON results saved to: results/qc/backtest_YYYYMMDD_HHMMSS.json

Error Handling

LEAN CLI not installed:

Error: LEAN CLI not installed
Install with: pip install lean
Then login with: lean login

Missing credentials:

Error: QC_PROJECT_NAME not set in .env
Add to .env file:
QC_PROJECT_NAME="Your Project Name"

Backtest timeout:

⚠ Backtest timed out after 300 seconds
Check status at: https://www.quantconnect.com/terminal

Integration with Claude Code

Auto-Activation Patterns

Automatically activate this skill when detecting:

• File changes in strategies/*.py or lean_projects/*/main.py
• Keywords: "backtest", "quantconnect", "test strategy", "qc cloud", "lean backtest"
• Strategy modification followed by validation requests
• Performance testing requests

Workflow Integration

1. Strategy Development:

   • Detect strategy file modifications
   • Execute bundled script with appropriate flags
   • Parse and present performance metrics

2. Result Analysis:

   • Use --wait --save flags to retrieve complete results
   • Parse JSON output for key metrics
   • Provide analysis and recommendations

3. Iterative Testing:

   • Modify strategy parameters
   • Commit changes if requested
   • Execute backtest and compare results

Success Criteria

Backtest Started: Exit code 0, "Backtest started" message Backtest Completed: Results JSON downloaded, metrics displayed Git Integration: Changes committed and pushed successfully Error Handling: Clear error messages with actionable steps

Related Skills

• data-manager: Download historical data before backtesting
• backtest-runner: Run local Backtrader backtests
• parameter-optimizer: Optimize strategy parameters before cloud testing

Notes

• GitHub Sync: If GitHub integration is enabled in QuantConnect, changes pushed to git@github.com:vinamra2013/QuantConnectAlgoTradingStratigies.git will auto-sync to QuantConnect within 1-2 minutes
• Rate Limits: QuantConnect Research Seat allows unlimited backtests
• Execution Time: Typical backtest (5 years daily data) completes in 30-90 seconds
• Cost: $27/month Research Seat (unlimited backtests)

Troubleshooting

Issue: "Project not found"

Solution: Check QC_PROJECT_NAME matches exact project name in QuantConnect terminal

Issue: "Authentication failed"

Solution: Run lean login and re-enter credentials from https://www.quantconnect.com/account

Issue: "Results not found"

Solution: Backtest may still be running. Increase --timeout or check web UI manually

Issue: Git push fails

Solution: Ensure SSH key is configured for git@github.com:vinamra2013/QuantConnectAlgoTradingStratigies.git