Performance and Advanced Settings

Fine-tune SystemSculpt for your device, debug issues, and unlock power-user features

Performance optimization (2 minutes)

Quick performance check

1. Enable monitor: Advanced → Show Performance Indicator ✓
2. See metrics: Status bar shows memory & lag
3. Too slow? Try these fixes:

Advanced settings overview

Developer tools

Debug mode - See everything:

Development Mode: ON
→ Console shows:
  - API calls & responses
  - Performance timing
  - Error details
  - Memory usage

When to enable:

• Troubleshooting issues
• Reporting bugs
• Optimizing performance
• Custom development

Embeddings performance

For large vaults (1000+ notes):

Processing tips:

• Run overnight for initial processing
• Close other apps during processing
• Use "Process" button manually for updates

MCP tools optimization

Tool permissions for speed:

Auto-approve safe tools = Faster workflows

Device-specific settings

Desktop optimization

High-performance setup:

Yaml
Copy to clipboard
Development Mode: OFF (unless debugging)
Performance Monitor: OFF (saves resources)
Embeddings Batch: 32
Default Model: Balance speed/quality

Resource-constrained:

Yaml
Copy to clipboard
Embeddings: Disabled or manual only
Default Model: Fast/efficient models
Auto-processing: OFF
Limit context window: 25%

Mobile settings

iOS/Android optimization:

- Embeddings: OFF
- Batch size: 4-8 max
- Auto-process: Never
- Font size: Medium/Large
- Simplified UI: ON

Why different?

• Limited memory (400-512MB)
• Battery preservation
• Thermal management
• Background limits

Platform detection

View your device info:

Advanced → Device Information
Shows:
- Platform & version
- Memory available
- CPU cores
- Screen size
- Browser capabilities

Use this to:

• Verify feature support
• Share with support
• Optimize settings
• Understand limits

Troubleshooting tools

Debug console

Access developer console:

• Windows/Linux: Ctrl + Shift + I
• Mac: Cmd + Option + I
• Filter by: [SystemSculpt]

What to look for:

Javascript
Copy to clipboard
[SystemSculpt] API call: POST /chat/completions
[SystemSculpt] Response time: 1.2s
[SystemSculpt] Memory usage: 245MB
[SystemSculpt] Error: Connection timeout

Export logs

For support tickets:

Advanced → Copy Logs for Support
→ Automatically redacts sensitive data
→ Includes last 1000 entries
→ Paste in Discord/email

Reset options

Three reset levels:

Hidden settings

In data.json

Location: .obsidian/plugins/systemsculpt-ai/data.json

Adjustable values:

Json
Copy to clipboard
{
  "contextWindowPercentage": 25,
  "maxVaultEntries": 300,
  "embeddingsBatchSize": 16,
  "performanceMonitoring": false,
  "debugLevel": "info"
}

Context window

What it controls: How much model context to use

• Default: 25%
• Higher = More context but higher cost
• Lower = Cheaper but less context

Adjust for:

• Long conversations: Increase to 50%
• Cost saving: Decrease to 10%
• Large contexts: Keep at 25%

Vault directory limits

maxVaultEntries: How many files to include in vault context

• Default: 300
• Large vaults: Reduce to 100
• Focused work: Increase to 500

Performance monitoring

Built-in metrics

Status bar shows:

• Memory: Current usage in MB
• Lag: Response delay in ms
• Updates: Every 3 seconds

What's normal?

• Memory: 200-500MB
• Lag: <100ms
• Spikes: During processing

Finding bottlenecks

Common culprits:

Performance tips

General optimization:

1. Close unused panes
2. Archive old chats
3. Limit plugin count
4. Use local models for tests
5. Clear cache monthly

Backup system

Automatic backups

What's backed up:

• All settings
• Custom providers
• API keys (encrypted)
• Favorites
• Directory configs

Schedule:

• Daily at midnight
• Last 7 kept
• Before major changes

Manual backup

Export settings:

Advanced → Backup Settings
→ Save to safe location
→ Name with date

Restore process:

Advanced → Restore Backup
→ Select backup file
→ Review changes
→ Confirm restore

Power user features

Custom integrations

Access plugin API:

Javascript
Copy to clipboard
// In console or custom plugin
const systemsculpt = app.plugins.plugins['systemsculpt-ai'];
const api = systemsculpt.api;

// Use API methods
await api.chat.send("Hello");

Event hooks

Available events:

• chat-created
• message-sent
• response-received
• model-changed
• file-processed

Batch operations

Process multiple files:

Javascript
Copy to clipboard
// Via console
const files = app.vault.getMarkdownFiles();
for (const file of files.slice(0, 10)) {
  await processWithAI(file);
}

Security considerations

API key safety

Best practices:

• Never commit to git
• Use environment variables
• Rotate regularly
• Monitor usage
• Revoke if compromised

Vault security

When using MCP tools:

• Review actions in chat
• Don't auto-approve writes
• Exclude sensitive folders
• Backup before bulk operations

Privacy settings

Minimize data sharing:

• Use local models when possible
• Disable telemetry
• Clear logs regularly
• Review provider policies

Common scenarios

Research setup

Yaml
Copy to clipboard
Embeddings: Enabled
Batch size: 16
Context window: 40%
Debug: OFF
Default model: Balanced performance model

Development setup

Yaml
Copy to clipboard
Debug mode: ON
Performance monitor: ON
All tools: Enabled
Console logging: Verbose
Default model: Advanced model

Privacy setup

Yaml
Copy to clipboard
Provider: Ollama only
Embeddings: Local (when available)
No cloud providers
No auto-processing
Logs: Disabled

Getting help

Diagnostic steps

1. Enable debug mode
2. Reproduce issue
3. Check console (Ctrl/Cmd + Shift + I)
4. Copy relevant logs
5. Disable debug mode

Support resources

• Discord: Real-time help
• GitHub: Bug reports
• Logs: Advanced → Copy for Support
• Device info: Share for platform issues

---

⚡ Remember: Default settings work for most users. Only tune if you have specific needs!