Skip to content

SECURITY.md

Latest commit

 

History

History
268 lines (197 loc) · 7.02 KB

SECURITY.md

File metadata and controls

268 lines (197 loc) · 7.02 KB

Security Policy

Supported Versions

Version Supported
1.0.x ✅
< 1.0 ❌

Security Model

Threat Profile

Lucid Terminal operates with a user-level threat model:

  • In Scope: Protection against malicious code execution, data exfiltration, privilege escalation within user context
  • Out of Scope: System-level attacks, kernel exploits, physical access attacks

Security Features

1. Code Execution Safety

LLM Output Sanitization:

  • All AI-generated code is syntax-validated before execution
  • No eval() or dynamic code execution in renderer process
  • Scripts run with user permissions only (no sudo/admin)
  • Sandboxed test environment for validation

Implementation:

// electron/core/executor/scriptExecutor.ts
async validateSyntax(code: string, language: string): Promise<boolean> {
  // Language-specific syntax validation
  // Returns false if malicious patterns detected
}

2. Data Protection

FixNet Encryption:

  • Algorithm: AES-256-GCM (NIST approved)
  • Key Derivation: PBKDF2 with 100,000 iterations
  • Authentication: HMAC-SHA256 for integrity
  • Storage: User home directory (~/.lucid/fixnet/)

No External Data Transmission:

  • 72% offline operation (no network calls)
  • Ollama API is local-only (localhost:11434)
  • No telemetry, analytics, or crash reporting
  • User data never leaves the machine

3. IPC Security

Electron Process Isolation:

// electron/preload.ts
contextBridge.exposeInMainWorld('lucidAPI', {
  // Only safe, validated methods exposed
  command: (input: string) => ipcRenderer.invoke('command', input),
  // No direct filesystem or shell access
});

Message Validation:

  • All IPC messages validated with TypeScript types
  • Input sanitization on both main and renderer sides
  • Rate limiting on expensive operations

4. Filesystem Access Control

Restricted Operations:

  • Scripts created in ~/Documents/lucid-scripts/ only
  • No write access to system directories
  • No automatic execution without validation
  • User confirmation for file operations

Permission Model:

  • Read: Current directory and subdirectories
  • Write: User home directory only
  • Execute: Validated scripts with user confirmation

5. Model Security

Ollama Sandboxing:

  • Models run in Ollama's sandboxed environment
  • No direct system access from models
  • Resource limits enforced by Ollama
  • No persistent state across calls

Model Validation:

  • Only officially published Ollama models supported
  • No custom model loading (prevents supply chain attacks)
  • Model checksums verified by Ollama

Known Limitations

Current Vulnerabilities

  1. Dependency Supply Chain

    • Risk: npm packages could be compromised
    • Mitigation: Regular npm audit, pinned versions
    • Status: Ongoing monitoring

  2. Electron Security

    • Risk: Electron framework vulnerabilities
    • Mitigation: Regular updates, nodeIntegration disabled
    • Status: Following Electron security best practices

  3. Local Privilege Escalation

    • Risk: User could execute malicious scripts
    • Mitigation: Syntax validation, sandboxed testing
    • Status: User responsibility for script review

Out of Scope

The following threats are not addressed by Lucid Terminal's security model:

  • Physical access attacks
  • Keyloggers or screen capture malware
  • OS-level privilege escalation
  • Side-channel attacks (timing, cache)
  • Social engineering attacks

Reporting a Vulnerability

Responsible Disclosure

We take security seriously. If you discover a vulnerability:

DO:

  • Report privately via email or GitHub Security Advisory
  • Provide detailed reproduction steps
  • Allow reasonable time for patch (45 days)
  • Coordinate disclosure timing with maintainers

DON'T:

  • Publicly disclose before patch is available
  • Exploit the vulnerability for personal gain
  • Test on systems you don't own

Reporting Process

1. Private Report (Preferred)

GitHub Security Advisory:

  1. Go to https://github.com/GareBear99/lucid-terminal/security/advisories
  2. Click "Report a vulnerability"
  3. Fill out the form with:
    • Vulnerability description
    • Affected versions
    • Reproduction steps
    • Proof of concept (if applicable)
    • Suggested fix (if known)

Email:

  • Contact: Via GitHub profile
  • Subject: [SECURITY] Lucid Terminal Vulnerability
  • Include: Same information as above

2. Public Report (Non-Critical)

For low-severity issues (documentation, UI bugs without security impact):

  • Open a regular GitHub issue
  • Tag with security label
  • No embargo required

Response Timeline

Severity Response Time Patch Time
Critical 24 hours 7 days
High 48 hours 14 days
Medium 5 days 30 days
Low 10 days 60 days

Severity Definitions:

  • Critical: Remote code execution, data exfiltration, privilege escalation
  • High: Authentication bypass, XSS, local code execution
  • Medium: Denial of service, information disclosure
  • Low: UI issues, non-exploitable bugs

Acknowledgments

Security researchers who responsibly disclose vulnerabilities will be:

  • Acknowledged in CHANGELOG.md (if desired)
  • Listed in SECURITY.md Hall of Fame
  • Invited to test patches before release

Hall of Fame:

  • No reports yet - be the first!

Security Best Practices for Users

1. Keep Updated

# Check for updates
gh release list --repo GareBear99/lucid-terminal

# Download latest
gh release download v1.0.0 --repo GareBear99/lucid-terminal

2. Review Generated Scripts

Before executing AI-generated code:

  • Read the script completely
  • Understand what it does
  • Check for suspicious operations (rm -rf, curl, wget)
  • Verify file paths and permissions

3. Use Least Privilege

Don't run Lucid Terminal as admin:

# ❌ Bad
sudo open "Lucid Terminal.app"

# ✅ Good
open "Lucid Terminal.app"

4. Isolate Sensitive Environments

For production systems:

  • Use separate user account for Lucid Terminal
  • Don't store credentials in scripts
  • Use environment variables for secrets
  • Enable audit logging

5. Monitor FixNet Dictionary

Regular audits:

# Check FixNet size
du -sh ~/.lucid/fixnet/

# Review recent additions
ls -lt ~/.lucid/fixnet/fixes/ | head

# Backup periodically
cp -r ~/.lucid/fixnet/ ~/.lucid/fixnet.backup/

Vulnerability History

v1.0.0 (2026-02-28)

  • Initial release
  • No known vulnerabilities

Security Contact

References

  1. Electron Security Checklist
  2. OWASP Top 10
  3. CWE Top 25
  4. NIST Cryptographic Standards

Last Updated: 2026-02-28
Version: 1.0.0