~/wiki

Remote Development Workflows

Confiance : high
remote-developmentsshvm-deploymentcursor-ideopenclawsclient-accessworkspace-isolationmulti-tenant-architecturesecurity-patternsdocker-containerizationper-client-deploymentgcp-deploymentshared-infrastructureclaude-code-integration

Development patterns and architectures for working on code deployed to remote servers, including considerations for multi-client environments and non-technical user access.

Traditional SSH-Based Development

Single Developer Model

Developer → SSH → Remote VM → Project Files → Deployment

Typical Setup:

  • GCP VM hosting multiple client projects
  • SSH access for code modification
  • cursor-ide with remote development capabilities
  • Direct file system access to all projects

Limitations:

  • Single point of access (developer bottleneck)
  • Shared environment security concerns
  • Manual coordination for multi-project management
  • No client self-service capabilities

Multi-Client Challenges

When hosting multiple client projects on shared infrastructure:

  • Cross-contamination risk: Accidental file access across projects
  • Security boundaries: SSH access exposes entire file system
  • Scaling bottleneck: Developer required for all changes
  • Cost inefficiency: Idle developer time for routine modifications

Client-Facing Remote Development

Architecture Evolution

Moving from developer-mediated to client-autonomous access:

Traditional:  Client → Developer → SSH → VM → Changes
Evolution:    Client → Bot Interface → Scoped Agent → VM → Changes  

OpenClaw Integration

openclaw enables secure client access to remote development through:

Workspace Isolation:

  • Container-based project separation
  • File system boundaries preventing cross-project access
  • Per-client deployment with dedicated resources

Conversational Interface:

  • Telegram/WhatsApp bots for natural language requests
  • No SSH knowledge or terminal access required
  • Claude API integration for intelligent code modification

Security Model:

  • Sandboxed execution environment per client
  • Approval workflows for sensitive operations
  • Audit trails for all modifications

Implementation Patterns

Per-Client Containerization

VM Host:
├── client-a/
│   ├── docker-compose.yml (OpenClaw + project)
│   ├── workspace/ (scoped file access)
│   └── config/ (client-specific settings)
├── client-b/
│   ├── docker-compose.yml
│   ├── workspace/
│   └── config/

Benefits:

  • Complete isolation between clients
  • Independent scaling and resource allocation
  • Simplified backup and migration strategies
  • Clear security boundaries

Hybrid Development Models

Developer + Client Access

  • Developer: Full SSH access for complex development
  • Client: Scoped bot access for content and minor modifications
  • Coordination: Shared deployment pipeline with appropriate permissions

Graduated Access Levels

Level 1: Content updates (text, images, configuration)
Level 2: Styling changes (CSS, layout modifications)  
Level 3: Feature modifications (functionality changes)
Level 4: System changes (dependencies, infrastructure)

Technology Stack Considerations

SSH-Based Development

Tools: Cursor, VS Code Remote, terminal access Pros: Full development environment, familiar tooling Cons: Technical barrier, security exposure, scaling limitations

Container-Based Development

Tools: Docker, OpenClaw, web-based interfaces Pros: Isolation, scalability, client accessibility Cons: Container overhead, complexity for simple projects

Cloud-Native Development

Tools: GitPod, CodeSpaces, cloud IDEs Pros: No local infrastructure, integrated CI/CD Cons: Vendor lock-in, cost for multiple clients, limited customization

Security Architecture

Network Isolation

  • VPN access: Secure tunnels for development access
  • Firewall rules: Port restrictions and IP whitelisting
  • Container networking: Isolated networks per client project

Authentication and Authorization

  • SSH key management: Separate keys per developer/client
  • Role-based access: Granular permissions by project area
  • Audit logging: Comprehensive tracking of all access and changes

Data Protection

  • Encrypted storage: At-rest encryption for client data
  • Backup strategies: Separate backup cycles per client
  • Compliance: GDPR, HIPAA, or other regulatory requirements

Deployment Strategies

Shared Infrastructure

Single VM hosting multiple client projects with logical separation:

Advantages: Cost efficiency, simplified management Disadvantages: Security concerns, resource contention, scaling limits

Dedicated Infrastructure

Separate VMs or cloud resources per client:

Advantages: Complete isolation, independent scaling, clear billing Disadvantages: Higher costs, management overhead, resource utilization

Hybrid Approach

Shared development infrastructure with isolated production deployments:

Development: Shared VM with container isolation Production: Dedicated resources for client applications Benefits: Development cost savings with production security

Client Onboarding

Technical Preparation

  1. Environment setup: Container deployment and configuration
  2. Bot creation: Dedicated communication channel (Telegram, etc.)
  3. AI configuration: Claude API setup with project context
  4. Testing: Validate functionality with sample requests

Client Training

  1. Interface introduction: How to communicate with the development bot
  2. Request patterns: Examples of effective change requests
  3. Limitation awareness: What the system can and cannot handle
  4. Escalation process: When to involve human developers

Monitoring and Maintenance

Performance Monitoring

  • Resource utilization: CPU, memory, disk usage per client
  • Response times: Bot response and deployment completion times
  • Error rates: Failed requests and system errors
  • Cost tracking: AI API usage and infrastructure costs

Quality Assurance

  • Automated testing: CI/CD pipelines with client-specific test suites
  • Code review: Optional developer oversight for complex changes
  • Rollback capabilities: Quick recovery from problematic deployments
  • Change logging: Audit trail of all modifications

See also