shancheas
ff6814d038
- Introduced coding standards for TypeScript and React in SKILL.md. - Added continuous learning skill with configuration and evaluation scripts. - Created detail layout guidelines for read-only pages. - Established form layout rules for data-entry forms. - Documented project guidelines for the frontend monorepo. - Implemented security review checklist for frontend/Electron applications. - Developed a verification loop skill for comprehensive session checks. This commit enhances the skill set available for developers, ensuring adherence to best practices and improving code quality.
53 lines
1.3 KiB
Plaintext
53 lines
1.3 KiB
Plaintext
---
|
|
description: Model selection, context window management, and build troubleshooting
|
|
alwaysApply: true
|
|
---
|
|
|
|
# Performance Optimization
|
|
|
|
## Model Selection Strategy
|
|
|
|
**Haiku 4.5** (90% of Sonnet capability, 3x cost savings):
|
|
- Lightweight agents with frequent invocation
|
|
- Pair programming and code generation
|
|
- Worker agents in multi-agent systems
|
|
|
|
**Sonnet 4.5** (Best coding model):
|
|
- Main development work
|
|
- Orchestrating multi-agent workflows
|
|
- Complex coding tasks
|
|
|
|
**Opus 4.5** (Deepest reasoning):
|
|
- Complex architectural decisions
|
|
- Maximum reasoning requirements
|
|
- Research and analysis tasks
|
|
|
|
## Context Window Management
|
|
|
|
Avoid last 20% of context window for:
|
|
- Large-scale refactoring
|
|
- Feature implementation spanning multiple files
|
|
- Debugging complex interactions
|
|
|
|
Lower context sensitivity tasks:
|
|
- Single-file edits
|
|
- Independent utility creation
|
|
- Documentation updates
|
|
- Simple bug fixes
|
|
|
|
## Ultrathink + Plan Mode
|
|
|
|
For complex tasks requiring deep reasoning:
|
|
1. Use `ultrathink` for enhanced thinking
|
|
2. Enable **Plan Mode** for structured approach
|
|
3. "Rev the engine" with multiple critique rounds
|
|
4. Use split role sub-agents for diverse analysis
|
|
|
|
## Build Troubleshooting
|
|
|
|
If build fails:
|
|
1. Use **build-error-resolver** agent
|
|
2. Analyze error messages
|
|
3. Fix incrementally
|
|
4. Verify after each fix
|