Pro Tips & Best Practices
Expert-level tips for maximizing productivity with Boring. From beginners to senior developers.
🎯 For Beginners
Start with Vibe Coding
You: /vibe_start Build a to-do app with React
Boring: Let me ask some questions first...
1. Do you want authentication?
2. Local storage or database?
3. Any specific styling framework?
Why it works: Vibe start guides you through requirements before coding.
Use Quick Fix Often
Automatically fixes all linting errors, import sorting, and formatting issues.Let AI Read Your Errors
You: I'm getting "ModuleNotFoundError: No module named 'foo'"
Boring: Encountered ❌ ModuleNotFoundError.
Running Active Recall...
Found pattern: "Missing dependency"
Solution: pip install foo
⚡ For Intermediate Developers
Tip 1: SpecKit Workflow
Before writing code, use the 5-step specification workflow:
graph LR
A[1. Constitution] --> B[2. Clarify]
B --> C[3. Plan]
C --> D[4. Checklist]
D --> E[5. Analyze]
"Measure Twice, Cut Once" - AI implementation!
Tip 2: Use Hybrid Mode
| Task Type | Recommended Approach |
|---|---|
| Small bug fixes | boring_apply_patch |
| New features | boring start + SpecKit |
| Code review | boring_code_review |
| Refactoring | boring_checkpoint + boring run |
Tip 3: Incremental Verification
# Only verify changed files (fast!)
boring verify --incremental
# Force full verification before release
boring verify --force --level FULL
Tip 4: Accumulate Knowledge
Develop → AI fixes errors → Recorded in .boring/memory
Project ends → boring_learn → Patterns saved to .boring/brain
Next project → AI auto-references past solutions!
🚀 For Senior Developers
Tip 5: Custom Quality Rules
Create .boring.toml with project-specific settings:
[boring.quality_gates]
min_coverage = 80 # High bar for production
max_complexity = 10 # Keep it simple
max_file_lines = 300 # Smaller files
[boring.linter_configs]
ruff_line_length = 100
ruff_ignore = ["E501", "W503"]
[boring.prompts]
judge_system = """
You are a senior Python architect focused on:
- SOLID principles
- Clean architecture
- Performance optimization
"""
Tip 6: Multi-Project RAG
boring_rag_search(
query="authentication middleware",
additional_roots=[
"/path/to/shared-libs",
"/path/to/other-project"
]
)
Search across multiple codebases simultaneously.
Tip 7: Contrastive Evaluation
# Compare two implementations
boring_evaluate(
level="PAIRWISE",
target_a="src/auth_v1.py",
target_b="src/auth_v2.py"
)
AI judges which implementation is better, with bias mitigation.
Tip 8: Security Sandbox with Shadow Mode
🧠 Cognitive Evolution (V11.2)
Tip 9: Levering Deep Thinking
Use "Deep Thinking" for architectural decisions or complex refactors.
Tip 10: Critical Thinking Audit
Use "Critical Thinking" for security reviews or logic verification.
🔧 Performance Tips
Tip 14: Warm the Cache
Tip 15: Parallel Verification
Tip 16: Selective RAG Indexing
🛡️ Security Tips
Tip 17: Always Use Shadow Mode in Production
Tip 18: Protected File Operations
# Use boring_write_file instead of native write_file
boring_write_file(path="config.py", content="...")
Tip 19: Secret Scanning
🔔 BoringDone Notifications
Tip 20: Completion Notifications (V12.2)
Get notified when AI tasks complete, so you don't have to watch the screen!
# AI automatically notifies you when a task is done
boring_done(task_name="Code Review", success=True)
Automatic Integrations:
- ✅ FlowEngine completion
- ✅ AgentLoop completion
- ✅ boring verify completion
See Also
- Quick Tutorials - Step-by-step guides
- MCP Tools - Tool reference
- Quality Gates - CI/CD setup