This is archived Verdent 1.0 documentation. For current product documentation, see Verdent.
Configuration Optimization
Optimize Verdent for speed, cost, and thoroughness
Optimization Areas
- Model selection for cost and performance balance
- Execution mode configuration for workflow efficiency
- Context management for large codebases
- Per-project and workspace-level customization
Speed vs Thoroughness Tradeoffs
Balance execution speed against analysis depth based on task requirements.
Configuration:
- Model: Claude Haiku 4.5 (fast, cost-efficient)
- Execution Mode: Agent Mode for direct execution
- Use Cases: Quick fixes, routine operations, simple tasks
Benefits:
- Fastest response times
- Minimal interruptions
- Lower credit consumption
Configuration:
- Model: Claude Sonnet 4.5 1M (extended context)
- Execution Mode: Plan Mode for strategic review
- Use Cases: Complex features, production code, architectural decisions
Benefits:
- Maximum context window (1M tokens)
- Complete visibility and control
- Review before execution
Most users benefit from context-switching between configurations:
- Quick fixes: Haiku + Agent Mode
- Complex features: Opus + Plan Mode
- Exploration: Plan Mode for strategy, then Agent Mode for execution
Strategy: Adapt configuration to task type instead of using one-size-fits-all approach.
Project-Specific Configuration
Project Rules (AGENTS.md)
- Location: Project root directory or workspace folder
- Scope: Applied only to the current project
- Version Control: Commit to git for team-wide standards
- Content: Coding standards, testing requirements, architectural patterns
Example:
# Project Rules
## Code Standards
- Use TypeScript strict mode
- Generate unit tests for all new functions
- Follow MVC pattern
## Testing
- All endpoints require input validation
- Run tests before committingConfiguration Priority
When configurations conflict, Verdent applies this priority order:
- Project Rules (AGENTS.md) - Highest priority (project-specific)
- Workspace Settings - Application workspace-level settings
- User Rules - Global user preferences (Settings → User Rules)
- Default Settings - Verdent's built-in defaults
Context Management for Large Projects
- Delegate complex operations to subagents with separate context windows
- Only subagent results consume main context, not entire process
- Prevents main context from filling with implementation details
- Explicitly add only relevant files using
@filename - Avoid loading entire project into context unnecessarily
- Target specific components for focused analysis
- Break large tasks into smaller, completable chunks
- Start new sessions for distinct features or components
- Natural breakpoints prevent context depletion
- Review approach before consuming execution context
- Reduces wasted context on incorrect implementations
- Plan Mode has separate context management
Multi-workspace scenarios automatically apply appropriate project rules when switching workspaces. No manual configuration switching required.
Desktop-Specific Optimization
Parallel Agent Efficiency
Verdent's parallel execution capability requires optimization consideration:
When Running Multiple Agents:
- Each agent maintains independent context
- CPU and memory usage scales with active agents
- Network requests parallelize for faster completion
Optimization:
- Limit concurrent agents on resource-constrained machines
- Use Claude Haiku 4.5 for routine parallel tasks
- Reserve Claude Sonnet 4.5 for complex individual tasks
Git Workspace Strategy:
- Each workspace provides full file isolation
- Agents in separate workspaces never conflict
- Base workspace serves as reference
Optimization:
- Group related tasks in the same workspace
- Create dedicated workspaces for long-running features
- Clean up completed workspaces to free resources
Performance Optimization
Model Selection Strategy
| Task Type | Recommended Model | Reason |
|---|---|---|
| Quick fixes | Claude Haiku 4.5 | Speed and cost |
| General development | Claude Sonnet 4.5 | Best tradeoff |
| Complex architecture | Claude Sonnet 4.5 1M | Extended context |
| Parallel routine tasks | Claude Haiku 4.5 | Resource efficiency |
Execution Mode Strategy
| Scenario | Recommended Mode | Reason |
|---|---|---|
| Learning new codebase | Plan Mode | Safe exploration |
| Active development | Agent Mode | Full capabilities |
| Production changes | Plan Mode first | Review before execution |
Cost Optimization
Credit Efficiency
Balance credit consumption against productivity:
Strategies:
- Use Claude Haiku 4.5 for routine tasks (lower credit consumption)
- Use
@verifierfor quick validation (separate context) - Break large tasks into smaller sessions
- Use Plan Mode to avoid wasted execution context
Impact: Significant credit reduction for typical workflows
Strategies:
- Reserve Claude Sonnet 4.5 1M for high-value complex tasks
- Use parallel agents for time-sensitive work
- Invest credits in thorough planning for production code
Impact: Higher per-credit productivity for complex work
Monitoring Credit Usage
Track consumption in the User Menu:
- Credit balance: Current credits remaining
- Subscription details: Plan information
- Top-up options: Purchase additional credits
FAQs
How do I optimize for large codebases?
Large codebase optimization strategies:
- Use subagents strategically:
@verifierpreserves main context - Read files selectively: Use
@filenamefor specific files only - Chunk large tasks: Break features into smaller, focused sessions
- Use Plan Mode first: Plan approach before consuming execution context
What's the best configuration for team collaboration?
Recommended team configuration:
- Project Rules (AGENTS.md): Commit shared coding standards
- Plan Mode for changes: Review approach before execution
- Auto Generate Code Review: Enable for quality assurance
This balances productivity with code review needs.
How do parallel agents affect performance?
Resource impact:
- CPU: Linear scaling with active agents
- Memory: Each agent maintains separate context
- Network: Parallel API requests
Optimization:
- 2-3 concurrent agents optimal for most machines
- Use Claude Haiku 4.5 for parallel tasks
- Monitor system resources during heavy parallel work