# ReadMe MCP

> ReadMe connects your AI client directly to your developer documentation hub. It turns your agent into an expert technical reader, allowing it to instantly find specific guides, API references, or changelogs without leaving the chat window. Your AI can search every page, track recent updates, and pull complex content from your entire knowledge base.

## Overview
- **Category:** developer-tools
- **Price:** Free
- **Tags:** api-documentation, developer-portal, technical-writing, knowledge-base, documentation-search, developer-experience

## Description

This MCP lets you connect your internal developer documentation directly to your agent. You no longer have to copy-paste snippets or switch between tabs just to find an API endpoint definition. Instead, your AI reads the source of truth—your published guides and technical specs—and answers questions using that data.

It handles everything from complex full-text searches across all categories to fetching the exact content for a specific changelog entry. You can even ask it to list all available documentation sections or analyze how different parts of your project are structured. This capability is critical because it keeps your agent grounded in *your* company's official information, not general web data. By connecting this MCP via Vinkius, you ensure that any compatible AI client—whether running locally or on a cloud platform—has immediate access to the full depth and breadth of your technical knowledge.

## Tools

### list_categories
Lists all major organizational groups within your entire set of documentation.

### get_category
Fetches detailed information about one specific section or category of guides.

### get_category_docs
Lists every single individual documentation page that belongs to a selected guide group.

### get_doc
Retrieves the full, raw content of any specific documented article or page.

### search_docs
Runs a comprehensive search across all articles and guides to find relevant information by keyword.

### list_changelogs
Pulls a list of the titles and dates for all published product update announcements.

### get_changelog
Retrieves the full text content for one specific, named changelog post.

### list_custom_pages
Shows a list of standalone pages that exist outside the main category structure.

### get_custom_page
Retrieves the full content for any specific custom, non-categorized page.

### get_project
Gathers metadata about the overall scope and configuration of your entire documentation project.

## Prompt Examples

**Prompt:** 
```
Search the documentation for instructions on configuring webhooks.
```

**Response:** 
```
Using the `search_docs` tool, I found several results. The top match is 'Webhook Installation Guide' (Slug: webhook-setup). It provides step-by-step instructions on setting endpoint URLs.
```

**Prompt:** 
```
Get the contents of the changelog titled 'v2-api-release'.
```

**Response:** 
```
I requested `get_changelog` for 'v2-api-release'. The post details new rate limits, updated endpoints for user management, and deprecation notices for the v1 metrics paths.
```

**Prompt:** 
```
List all main documentation categories.
```

**Response:** 
```
I fetched the configuration using `list_categories`. There are 4 global categories: 'Getting Started', 'Authentication', 'REST API Reference', and 'Webhooks Guide'.
```

## Capabilities

### Search all documentation
The agent performs full-text searches across every guide, reference page, and custom document.

### Retrieve specific content pages
You can fetch the complete Markdown text for any single documented page or API reference.

### List documentation structure
The agent maps out your entire knowledge base, showing all major categories and sub-guides available.

### Track product changes
You can pull recent updates, new features, or deprecation notices directly from published changelogs.

### Analyze project structure
The agent provides details on the overall organization and scope of your entire ReadMe project.

## Use Cases

### The engineer needs API specs for a new feature.
Instead of searching Google or asking a colleague, the agent uses `search_docs` and retrieves the exact current endpoint requirements. This lets the software engineer validate their code against official standards in minutes.

### The developer advocate needs to create an 'Authentication' guide.
They ask the agent to list all related documentation (`get_category_docs` for Authentication). The AI provides a comprehensive map, ensuring no necessary sub-guide is missed before writing begins.

### The technical writer needs to summarize the latest v2 API changes.
They run `list_changelogs` first, identifying 'v2-api-release'. Then they use `get_changelog` to pull the full content, allowing them to generate a summary for internal stakeholders immediately.

### The team needs to know what's in the project.
They use `list_custom_pages` and `get_project` together. The agent confirms that key standalone guides exist, providing an overview of content outside the main structured categories.

## Benefits

- Instantly check product status: Instead of hunting through multiple sources, use the `list_changelogs` tool to see a list of all recent updates and announcements immediately.
- Audit content accuracy: If you need to verify an old reference, calling `get_doc` lets your agent pull the full text of that specific page for review, without needing manual logins or searches.
- Understand structure quickly: Use `list_categories` to map out your entire documentation hierarchy. Your agent reports back on all major sections available in a single response.
- Write targeted content: If you need details on one topic but the guide is nested deep, use `get_category_docs` to list all relevant sub-pages before drafting.

## How It Works

The bottom line is that you get reliable, real-time access to your entire technical knowledge base from a single conversation thread.

1. Authorize the ReadMe MCP in your environment and provide your API key.
2. Instruct your AI client to perform a task, like summarizing an API endpoint or listing all categories.
3. The agent uses the tool to read the documentation and responds with accurate, source-backed information.

## Frequently Asked Questions

**How does ReadMe MCP help me find a specific API endpoint?**
You use the `search_docs` tool. Instead of guessing keywords, your agent performs a full-text search across all guides and returns direct links and content snippets matching the endpoint you need.

**Can I list everything in my documentation?**
Yes, start by using `list_categories` to see the main sections. Then, use `get_category_docs` on any category name to get a full manifest of all related pages.

**How do I know what was changed recently?**
Run the `list_changelogs` tool first. This gives you a list of titles and dates for all updates, helping you pinpoint exactly which version you need to read about next with `get_changelog`.

**Does ReadMe MCP work only on guides?**
No, it's more comprehensive. Besides structured guides, you can also check standalone content by using `list_custom_pages` and fetching the details with `get_custom_page`.

**What if I need to summarize a whole guide?**
You should use `get_doc` on the specific page you want summarized. While the tool fetches the content, your agent can then process that full text into the summary you request.