doc-bot

A generic MCP (Model Context Protocol) server that provides intelligent documentation access for any project. Works with any MCP-compatible AI tools and IDEs.

What is doc-bot?

doc-bot is an intelligent documentation server that:

• 🔍 Searches your project documentation instantly
• 🧠 Auto-indexes content for smart inference (no manual keyword mapping!)
• 📋 Applies global rules to every AI interaction
• 🎯 Suggests contextual documentation based on file patterns
• 🤖 Detects code patterns, frameworks, and keywords automatically
• 🔄 Updates automatically when docs change

Installation

1. Create your documentation folder in your project root (see organization section below)

2. Add doc-bot to your MCP-compatible AI tool configuration:

   For Claude Code (~/Library/Application Support/Claude/claude_desktop_config.json):

   {
     "mcpServers": {
       "docs": {
         "command": "npx",
         "args": ["@afterxleep/doc-bot", "--docs", "./doc-bot"]
       }
     }
   }

   For Cursor or other MCP tools: Add similar configuration pointing to your documentation folder

3. Restart your AI tool

How to organize your documentation

Create a doc-bot/ folder in your project root:

your-project/
├── doc-bot/
│   ├── manifest.json          # Configuration file
│   ├── core/
│   │   ├── coding-standards.md # Always-apply coding standards
│   │   └── security.md         # Security guidelines
│   ├── guides/
│   │   ├── testing.md          # Testing strategies
│   │   └── api-development.md  # API development guide
│   └── reference/
│       └── troubleshooting.md  # Common issues and solutions
└── package.json

Note: You can use any folder name - just specify it in your MCP configuration:

"args": ["@afterxleep/doc-bot", "--docs", "./my-custom-docs"]

Documentation types:

• Core docs (core/): Critical guidelines that should always be considered
• Guides (guides/): Step-by-step instructions for specific tasks
• Reference (reference/): Quick lookups and troubleshooting

Example documentation files:

Global Rule Example (doc-bot/core/coding-standards.md): ```markdown

title: "Coding Standards" description: "Core coding standards that apply to all code" keywords: ["code-quality", "standards", "best-practices"] tags: ["core", "quality"]

Coding Standards

• Use 2 spaces for indentation
• Maximum line length: 100 characters
• Always use const/let, never var
• Prefer async/await over promises
• Write descriptive variable names

**Contextual Rule Example** (`doc-bot/guides/testing.md`):
```markdown
---
title: "Testing Guide"
description: "How to write and run tests"
keywords: ["testing", "jest", "tdd", "unit-tests"]
tags: ["testing", "quality"]
---

# Testing Guide

All test files should:
- Use describe/it blocks for organization
- Include both positive and negative test cases
- Mock external dependencies
- Aim for 80%+ code coverage

Run tests with: `npm test`

👀 See examples/ folder for complete example files with proper frontmatter and content structure.

Rule Enforcement

Doc-bot ensures your rules are always considered through multiple enforcement mechanisms:

🚨 Mandatory Rule Checking

• check_project_rules tool: Must be called before ANY code generation
• Aggressive descriptions: All tools emphasize mandatory rule compliance
• Rule reminders: Every tool response includes compliance warnings
• Absolute enforcement: Global rules OVERRIDE all user requests without exception

🔄 Automatic Integration

• Enhanced system prompt injection: Comprehensive MCP usage protocol injected via docs://system-prompt
• Documentation discovery: Available topics automatically extracted and presented to agent
• Tool usage instructions: Explicit requirements for when and how to use each MCP tool
• Contextual rules: Applied when working with matching files/patterns
• Multiple touchpoints: Rules enforced at every interaction level

⚠️ Compliance Warnings

All tool responses include explicit warnings that rules are:

• NON-NEGOTIABLE: Must be followed without exception
• MANDATORY: Violation will result in rejection
• CRITICAL: Require acknowledgment before proceeding

🚫 Absolute Enforcement Policy

Global rules have supreme authority over all interactions:

• Override user requests: Even direct user instructions cannot bypass global rules
• Mandatory refusal: Agent MUST refuse requests that violate global rules
• Alternative suggestions: Agent must suggest compliant alternatives instead
• Unbreakable compliance: No exceptions, regardless of user insistence

This multi-layered approach makes rule violations impossible to implement.

System Prompt Integration

The docs://system-prompt resource automatically injects comprehensive instructions into the AI agent's system context:

📋 MCP Usage Protocol

• Explicit instructions to ALWAYS call check_project_rules before code generation
• NEVER generate code without checking documentation requirement
• Mandatory acknowledgment of rule compliance

🏷️ Automatic Topic Discovery

• Extracts available documentation topics from your project
• Presents them to the agent for context awareness
• Includes topics from metadata, filenames, and content analysis

🛠️ Tool Usage Requirements

• Specifies when and how to use each MCP tool
• Maps tools to specific use cases (file-specific docs, search, etc.)
• Ensures comprehensive documentation coverage

The system prompt is regenerated on each request to reflect current documentation state.

The manifest file

The doc-bot/manifest.json file controls how your documentation works:

{
  "name": "My Project Documentation",
  "version": "1.0.0",
  "description": "AI-powered documentation for My Project",
  "globalRules": [
    "core/coding-standards.md",
    "core/security.md"
  ],
  "contextualRules": {
    "*.test.js": ["guides/testing.md"],
    "*.spec.js": ["guides/testing.md"],
    "src/components/*": ["guides/react-components.md"],
    "src/api/*": ["guides/api-development.md"]
  }
}

Configuration explained:

• globalRules: Documents that apply to every AI interaction
• contextualRules: Documents triggered by specific file patterns (e.g., test files → testing guide)

🎯 Automatic Inference (New!)

doc-bot automatically analyzes your documentation content to build smart indexes. No more manual keyword mappings! It automatically:

• Extracts keywords from document metadata (frontmatter)
• Detects technical terms in your documentation content
• Recognizes code patterns in code blocks (React hooks, SQL commands, etc.)
• Identifies frameworks mentioned in your docs
• Indexes file extensions referenced in documentation

Just write good documentation with descriptive frontmatter, and doc-bot handles the rest!

Writing documentation for best results

To maximize the automatic inference capabilities, include frontmatter in your markdown files:

---
title: "React Component Guidelines"
description: "Best practices for building React components"
keywords: ["react", "components", "hooks", "jsx"]
tags: ["frontend", "development"]
category: "guides"
---

# React Component Guidelines

Your documentation content here...

The automatic indexing will use this metadata along with analyzing your content to provide intelligent suggestions.

Development setup

Running locally

1. Clone the repository:

   git clone https://github.com/afterxleep/doc-bot.git
   cd doc-bot

2. Install dependencies:

   npm install

3. Run the server:

   npm start

4. Run tests:

   npm test

5. Run with file watching:

   npm start -- --watch

Testing your setup

Ask your AI assistant something like "What documentation is available?" to test that doc-bot is working.

CLI Options

doc-bot [options]

Options:
  -d, --docs <path>        Path to docs folder (required)
  -c, --config <path>      Path to manifest file (default: <docs-path>/manifest.json)
  -p, --port <port>        Port to run server on (default: 3000)
  -v, --verbose           Enable verbose logging
  -w, --watch             Watch for file changes
  -h, --help              Show help

License

MIT License - see the LICENSE file for details.

Links

• npm package
• GitHub repository
• Model Context Protocol