- STRUCTURED
- LOGIC
- SCHEMA
- HUMAN
Short answer
How can I get my AI to understand my API documentation?
DocBreach pulls the technical details from your OpenAPI, Swagger, or Postman files and turns them into structured data. Your AI uses this data to identify available endpoints, required parameters, and service logic. You get a clear map of your service capabilities.
Where the data lands.
The outcomes of extraction.
The AI processes your documentation into these specific formats.
STRUCTURED
Clean endpoint maps
The AI uses docs_extract to pull paths and methods into a format it can actually reason about. You get a reliable list of what's available.
LOGIC
Service capability maps
Your agent understands the relationship between different endpoints. This lets it plan complex sequences of API calls.
SCHEMA
Parameter definitions
The AI identifies exactly which headers, query params, and body objects are required. This prevents errors when your agent tries to call an endpoint.
HUMAN
Manual review needed
The AI identifies the endpoints, but you decide which ones are safe for the agent to execute. You maintain control over the actual API calls.
The workflow
What your AI does when a collection arrives.
The AI handles the heavy lifting of parsing messy documentation files.
Ingest documentation
Your agent reads the raw OpenAPI, Swagger, or Postman files you provide.
docs_extractParse endpoints
The AI identifies every unique URL path and the HTTP methods associated with them.
docs_extractMap parameters
It extracts the specific data types and required fields for every request.
docs_extractBuild capability model
The agent organizes the extracted data into a logical map of what the service can actually perform.
docs_extract
Try it
Copy these to start.
Use these prompts once you've connected DocBreach.
Starting points
These are opening lines. Swap out the bracketed text with your specific file names or service names.
DocBreach Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_irgtI6t1CMNMo7SoPiRQBNgk7Rz1GQFQBkyR1OUW/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — DocBreach capabilities are ready to use.
{
"mcpServers": {
"docbreach-mcp": {
"url": "https://edge.vinkius.com/vk_preview_irgtI6t1CMNMo7SoPiRQBNgk7Rz1GQFQBkyR1OUW/mcp"
}
}
}Extract all available endpoints from my [filename.json] file.
Based on this Swagger doc, what are the required parameters for the [endpoint name] call?
Map out the service capabilities described in this Postman collection.
Summarize the authentication requirements for the APIs in [filename.yaml].
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect DocBreach once, then ask.
Just click the link to connect the MCP to your client. Your credentials stay encrypted, and you can immediately start asking your agent about your API docs.
Connect DocBreach to your AIFAQ
How this task behaves.
- 01
Can the AI automatically call these APIs for me?
No. DocBreach extracts the documentation so your AI understands the endpoints, but the AI does not execute the calls itself.
- 02
Does it work with Postman collections?
Yes, it can parse Postman collections into structured data.
- 03
Will the AI change my original Swagger files?
No. The AI only reads the files to extract information. It cannot modify or overwrite your source documentation.
- 04
Can it handle large OpenAPI files?
Yes, the extraction process is designed to handle complex and large-scale service definitions.
- 05
Does the AI decide which endpoints are safe to use?
No. The AI provides the map of what is possible, but you must decide which endpoints the agent is allowed to interact with.
More questions about DocBreach? The Connector page answers them. See everything the DocBreach Connector can do
Connect DocBreach to Claude, Cursor, ChatGPT & more
