Specific compliance certifications and detailed security policies are maintained by the Verdent team. Contact support@verdent.ai for compliance documentation, security questionnaires, or enterprise security requirements.
Your Code’s Privacy
Code Processing
How Verdent processes your code:- Code is transmitted to AI model servers for processing
- Processing occurs in real-time during active sessions
- No permanent storage of your code beyond session context
- Files you explicitly work with during conversations
- Project files read through file tools (file_read, glob, grep_content, grep_file)
- Command outputs from bash tool executions
- Chat conversation history within the session
- Files not opened or referenced during conversations
- Other VS Code windows or workspaces
- System files outside VS Code workspace
- Personal data unrelated to coding tasks
Data Transmission
Secure Communication:- All data transmitted over encrypted connections (HTTPS/TLS)
- Communication between VS Code extension and AI servers is encrypted
- No third-party access to transmission channels
Data Handling Policy
Data Collection
What data Verdent collects:- Code snippets processed during active sessions
- Chat conversation history
- File operations performed (read, edit, write)
- Tool usage patterns and commands executed
- Error logs and diagnostic information
- Provide AI-powered coding assistance
- Improve model responses and accuracy
- Debug technical issues
- Enhance user experience
Data Retention
Specific data retention policies, including how long conversation history and code snippets are retained, are managed by Verdent. Contact support@verdent.ai for detailed retention policy documentation.
- Active conversation context maintained during session
- Context cleared when starting new conversation
- Subagent contexts isolated and cleared after task completion
Data Sharing
Third-party sharing: Verdent processes code using AI model providers (Anthropic, OpenAI, Google). Your code is transmitted to these providers’ servers for processing. What is not shared:- Code is not sold to third parties
- No advertising or marketing use of your code
- No public disclosure of proprietary code
Security Best Practices
- Safe Usage
- MCP Security
- Bash Safety
Safe Usage Practices
1. Use Manual Accept Mode for sensitive projects: Click “Switch Permission” button in Input Box → Select “Manual Accept Mode”Review all file operations and commands before approval.2. Never include secrets in code:- Avoid hard-coded API keys, passwords, or credentials
- Use environment variables for sensitive configuration
- Review bash commands for credential exposure
- Verdent operates within current VS Code workspace only
- Sensitive projects should be in separate workspaces
- Use project-specific AGENTS.md for security rules
Compliance & Certifications
Industry Standards
Verdent’s compliance status, certifications, and audit reports are maintained by the security team. For specific compliance documentation (SOC 2, GDPR, HIPAA), security questionnaires, or enterprise requirements, contact support@verdent.ai.
- SOC 2: Security controls and practices
- GDPR: European data protection regulations
- HIPAA: Healthcare data requirements (if applicable)
- ISO 27001: Information security management
Regional Considerations
Data Processing Locations: AI model servers may process data in various regions. For region-specific data residency requirements, contact support@verdent.ai. GDPR Rights: European users have rights under GDPR including:- Right to access personal data
- Right to rectification
- Right to erasure
- Right to data portability
Submitting a Security Vulnerability Report
Responsible Disclosure
If you discover a security vulnerability:1
Do not disclose publicly
- Contact security team privately first
- Allow time for patches before public disclosure
- Responsible disclosure protects all users
2
Report to security team
- Email: support@verdent.ai
- Subject: “Security Vulnerability Report”
- Mark as urgent/confidential
3
Include in your report
- Vulnerability description and reproduction steps
- Potential impact and severity assessment
- Your contact information and suggested fixes (optional)
4
Expected response timeline
- Initial acknowledgment: Within 48 hours
- Status updates: As investigation progresses
- Resolution timeline: Based on severity
Security Resources
Staying Informed
Security Updates:- Check VS Code marketplace for security-related release notes
- Enable auto-update for VS Code extensions
- Monitor Verdent communications for security announcements
- Tool Reference - Bash command execution
- Best Practices - Safe prompting techniques
- Execution Modes - Manual Accept Mode