跳轉至

🧠 External Intelligence

Make Boring not just code, but possess "Real-time Knowledge" and "Deep Thinking" capabilities.

Boring integrates the most powerful MCP tools by default, giving your AI Agent superpowers.


📚 Context7: Your Real-time Documentation Library

Pain Point: AI models have training data cutoffs and don't know the latest API changes (e.g., Next.js 14, Stripe API updates). Solution: Context7 is a real-time RAG service specifically for technical documentation.

✨ Key Features

  • Real-time Queries: context7_query_docs fetches the latest library usage.
  • Precise Parsing: Automatically finds the most relevant Code Snippets.
  • Seamless Integration: Boring's boring-route determines when external docs are needed.

💡 Use Case

When you ask: "How to use the latest LangChain v0.2 LCEL?" 1. Agent detects training data might be stale. 2. Automatically calls context7 to query LangChain v0.2 docs. 3. Generates correct code based on the latest docs, not old patterns.


🤔 Thinking Tools: Deep Thinking Engine

Boring integrates two powerful thinking modes, forcing the Agent to "think clearly" before coding.

1. Sequential Thinking

Use Case: Complex refactoring, architecture design, multi-step tasks.

It forces the Agent into a "Chain of Thought": - Step 1: Analyze status - Step 2: Formulate hypothesis - Step 3: Verify hypothesis... - Only starts coding after thinking is complete.

2. Critical Thinking

Use Case: Code Review, bug hunting, security audit.

It teaches the Agent "Self-Doubt": - "Does this solution have side effects?" - "Is there a more efficient way?" - "Are edge cases considered?"


🛠️ Zero Config: One-Click Integration

You DO NOT need to manually download or install these external tools!

Boring uses an "All-in-One" design. Core features are built-in, and external intelligence is managed via a one-click extension mechanism.

How to Enable?

  1. Initialize Extensions (Run once):

    # In CLI
    boring setup-extensions
    
    Or ask the Agent in chat: > "Help me setup extensions" > (Agent will call boring_setup_extensions() tool)

  2. Auto-Configuration: Boring will automatically:

    • Download necessary MCP Servers (context7, criticalthink, sequential-thinking)
    • Configure API Keys and environment variables
    • Update configuration to enable them
  3. Ready to Use: Once setup, these tools appear automatically in the Agent's toolbox. You can ask directly in chat: > "Please use Sequential Thinking to plan this refactor." > "Please use Context7 to check the new React 19 Hooks."

This way, you get not just code, but high-quality output that has been "thought through" and "verified".