Repository Import Guide
Import your existing GitHub repositories to create intelligent, context-aware build plans. This guide will help you successfully import repositories and troubleshoot common issues.
Overview
Repository Import allows you to:
- Connect your GitHub account with secure OAuth
- Select and scan repositories for automated analysis
- Generate context-aware build plans that understand your existing codebase
- Continue conversations with persistent state across page reloads
Step-by-Step Import Process
1. Start Repository Import
- Navigate to Build Plans in your dashboard
- Click "Create New Project"
- Select "Import from Repository"
- You'll be redirected to the import wizard
2. Connect GitHub Account
First Time Setup:
- Click "Connect GitHub"
- You'll be redirected to GitHub OAuth
- Review and approve the permissions:
- ✅ Read access to public and private repositories
- ✅ User profile information
- ❌ No write access - we can't modify your code
- You'll be redirected back to Buildable with a success message
Permission Details:
- We only request read-only access
- We can't modify, delete, or push to your repositories
- You can revoke access anytime in your GitHub settings
- We don't store your source code, only project metadata
3. Select Repository
- Choose repository type using the dropdown filter:
- All Repositories - Shows both personal and organization repos (default)
- Personal Only - Only repositories you own
- Organizations - Only organization repositories you're a member of
- Browse your repositories in the list
- Use search to find specific repositories by name, description, or language
- Sort options:
- Recent - Recently updated repositories (default)
- Name - Alphabetical order
- Created - Recently created repositories
Repository Requirements:
- Must be accessible with your GitHub permissions
- Maximum size: 100MB (for optimal scanning performance)
- Supports all major programming languages
Organization Repository Access:
- Repositories from organizations you're a member of are now included
- Organization name is displayed as a badge next to repository name
- Same scanning and analysis capabilities as personal repositories
4. Repository Scanning
Once you select a repository:
- Scanning starts automatically - No additional action needed
- Real-time progress shown with scanning status
- Analysis includes:
- 📁 File structure and organization patterns
- 🔧 Technology stack detection (languages, frameworks, databases)
- ✨ Existing features identification
- 🏗️ Architecture patterns analysis
- 📊 Project statistics (file counts, sizes, complexity)
Scanning Timeline:
- Small projects (< 50 files): 10-30 seconds
- Medium projects (50-200 files): 30-60 seconds
- Large projects (200+ files): 1-3 minutes
5. Context-Aware Conversation
After scanning completes:
- Automatic navigation to conversation page
- Pre-populated project details from repository analysis
- Intelligent initial message that references your specific codebase
- Ready to plan enhancements and new features
Understanding Scan Results
Technology Detection
The scanner identifies:
Languages & Frameworks:
- Primary language (TypeScript, JavaScript, Python, etc.)
- Framework detection (React, Next.js, Django, Express, etc.)
- Build tools (Webpack, Vite, etc.)
Dependencies & Infrastructure:
- Package.json analysis for Node.js projects
- Requirements.txt for Python projects
- Database technologies (PostgreSQL, MongoDB, etc.)
- Deployment configurations (Docker, Vercel, etc.)
Example Scan Summary:
Technologies Found: TypeScript, React, Next.js, Supabase
Features Detected: Authentication, User Management, Dashboard
Project Structure: Component-based architecture, API routes
File Types: 45 .ts files, 29 .tsx files, 3 .css files
Feature Recognition
Our AI identifies existing features by analyzing:
- Component names and structure
- API endpoints and routes
- Database schema patterns
- Configuration files
- Documentation and comments
Common Feature Categories:
- 🔐 Authentication - Login, signup, session management
- 👥 User Management - Profiles, settings, permissions
- 📊 Analytics - Dashboards, metrics, reporting
- 💰 Payments - Billing, subscriptions, transactions
- 📧 Communication - Email, notifications, messaging
- 🛠️ Admin Tools - Management interfaces, configurations
Architecture Analysis
The scanner understands:
- Component organization patterns
- State management approaches
- API design conventions
- Data flow architecture
- Testing strategies
- Error handling patterns
Troubleshooting Common Issues
GitHub Connection Issues
Problem: "GitHub connection failed" error
Possible Causes & Solutions:
-
OAuth popup blocked
- Enable popups for bldbl.dev in your browser
- Try again with popup blocker disabled
-
Permission denied during OAuth
- Ensure you click "Authorize" on the GitHub page
- Check if your organization requires admin approval
- Try connecting with a personal repository first
-
Network or timeout issues
- Check your internet connection
- Try refreshing the page and reconnecting
- Clear browser cache and cookies
Verification Steps:
# Check if connection worked
1. Look for "GitHub Connected" success message
2. Verify you can see repositories in the list
3. Check for your username in the top-right corner
Repository Scanning Problems
Problem: Scan gets stuck in "Scanning..." status
Possible Causes & Solutions:
-
Large repository size
- Repository may exceed 100MB limit
- Try scanning a smaller repository first
- Contact support for enterprise scanning options
-
Network timeout
- Refresh the page and try again
- Check internet connection stability
- Wait a few minutes and retry
-
Repository access issues
- Verify you have read access to the repository
- Check if repository is private and permissions are correct
- Try with a public repository to test
Retry Process:
- Click the "Retry Scan" button if available
- Or go back and select the repository again
- Wait for the full scanning process to complete
Problem: "No files found" or incomplete scan results
Possible Causes & Solutions:
-
Empty or minimal repository
- Repository has very few files to analyze
- Try with a repository that has more substantial code
-
Unsupported file types
- Scanner focuses on code files (.js, .ts, .py, etc.)
- Documentation-only repositories may have limited results
-
Access restrictions
- Some files may be inaccessible due to GitHub permissions
- Try with a repository you fully own
Conversation Issues
Problem: Conversation starts from scratch after page reload
Solutions:
-
Check URL parameters
- Ensure the
snapshot=xyz
parameter is in the URL - If missing, go back to repository selection
- Ensure the
-
Clear browser data
- Clear localStorage for bldbl.dev
- Refresh and try the import process again
-
Repository context mismatch
- If you switched repositories, conversation resets automatically
- This is expected behavior for security
Problem: Repository context not showing in conversation
Solutions:
-
Verify scan completion
- Ensure scanning finished successfully
- Check that scan results show technologies and features
-
Refresh conversation page
- Reload the page to fetch latest repository data
- Look for the blue "Repository context loaded" indicator
-
Check snapshot validity
- Scan data may have expired (7 days)
- Re-scan the repository if needed
Permission and Access Issues
Problem: "Repository not found" or "Access denied" errors
Solutions:
-
Verify repository access
- Check you can access the repository on GitHub
- Ensure repository still exists and wasn't deleted
-
Update GitHub permissions
- Go to GitHub Settings → Applications → Buildable
- Ensure proper repository access is granted
-
Organization repositories
- Organization admin may need to approve access
- Try with personal repositories first
Problem: "GitHub token expired" error
Solutions:
-
Reconnect GitHub account
- Go through the GitHub connection process again
- This will refresh your access token
-
Check token expiration
- GitHub tokens expire after a certain period
- Regular reconnection may be required
Best Practices
Choosing Repositories for Import
Ideal Repositories:
- 🎯 Active projects you're currently enhancing
- 📊 Substantial codebases (50+ files) for better analysis
- 🏗️ Well-structured projects with clear patterns
- 🔧 Modern tech stacks for best AI understanding
Less Ideal:
- 📚 Documentation-only repositories
- 🗄️ Legacy code without clear structure
- 🔬 Experimental or prototype repositories
- 📦 Fork repositories without significant changes
Optimizing Scan Results
Repository Preparation:
- Update README with current project description
- Ensure package.json (for Node.js) or equivalent is up to date
- Clean up unused files to improve signal-to-noise ratio
- Use descriptive commit messages for better context
For Better Feature Detection:
- Use clear component and file names
- Include API documentation or comments
- Maintain consistent project structure
- Keep dependencies updated
Conversation Tips
Starting the Conversation:
- ✅ "I want to add user roles to my existing auth system"
- ✅ "Help me plan a notification feature for my Next.js app"
- ✅ "What's the best way to add payments to my SaaS?"
Avoid Generic Questions:
- ❌ "Help me code"
- ❌ "What should I build?"
- ❌ "Fix my app"
Reference Your Codebase:
- "Based on my existing user management..."
- "Given my current API structure..."
- "To integrate with my Supabase setup..."
Advanced Features
Multi-Repository Planning
Coming Soon:
- Import multiple related repositories
- Understand microservice dependencies
- Plan features across service boundaries
Incremental Updates
Future Enhancement:
- Webhook integration for automatic updates
- Detection of new features added manually
- Smart re-scanning of changed files
Team Collaboration
Planned Features:
- Share imported repository context with team members
- Collaborative planning sessions
- Team-wide repository access management
Getting Help
Self-Service Resources
- API Documentation - Technical integration details
- Integration Guides - Integrations overview
- FAQ - Common questions and answers
Contact Support
When to Contact Support:
- Repository import fails repeatedly
- Large repository scanning needs (>100MB)
- Organization-specific permission issues
- Enterprise deployment questions
How to Contact:
- 📧 Email: support@bldbl.dev
- 💬 Discord: Join our community
- 🐛 Bug Reports: GitHub Issues
Information to Include:
- Repository URL (if public) or name/size if private
- Error messages or screenshots
- Browser and version
- Steps to reproduce the issue
Quick Reference
Import Checklist
- GitHub account connected successfully
- Repository appears in the list
- Repository size is under 100MB
- Scan completes with technology detection
- Conversation loads with repository context
- Can ask questions about existing codebase
Common URLs
- Import Start:
/onboarding
- Repository Selection:
/onboarding/import
- Conversation:
/onboarding/conversation?snapshot={id}
- Dashboard:
/dashboard
Scan Status Meanings
pending
- Scan queued, will start shortlyscanning
- Actively analyzing repositorycompleted
- Successfully finished, ready for conversationfailed
- Error occurred, check error message and retry
Ready to import your first repository? Start here →