Appearance

Using Treezor documentation with AI assistants ​

Treezor publishes documentation in formats that AI assistants and development agents can retrieve directly. You can connect through the MCP server for real-time search, or use text exports that agents can search and parse with standard tools like grep or ripgrep.

The MCP server suits interactive use, while text exports suit agents that process files locally. Both approaches are valid and can be combined.

Choose the right source ​

NeedRecommended source
Search documentation from your AI toolMCP server
Discover relevant documentation/llms.txt
Read one topicPer-page Markdown (use the Download as Markdown button)
Verify methods, parameters, enums, schemasOpenAPI YAML files
Explore the API interactivelyScalar viewers in API References
Search or index the full documentation locally/llms-full.txt

MCP server ​

Treezor provides an MCP server that lets AI assistants search the documentation directly. The Model Context Protocol (MCP) is an open standard for connecting AI applications to external data sources and tools.

Server URL:

https://PDFXG05505.algolia.net/mcp/1/gvhzAVSBSOWJIKo8HQLrRw/mcp

The server is public and requires no authentication.

Setup ​

MCP is supported by Claude Code, Codex, Cursor, VS Code (with GitHub Copilot), and other AI development tools. The setup varies by tool.

Add the server using the Claude Code CLI:

bash
claude mcp add --transport http treezor-docs https://PDFXG05505.algolia.net/mcp/1/gvhzAVSBSOWJIKo8HQLrRw/mcp

Verify the connection:

bash
claude mcp list

To make the server available across all your projects, add --scope user to the add command.

For detailed setup instructions, see the Claude Code MCP documentation.

What the MCP server provides ​

The MCP server exposes search tools that let your AI assistant query Treezor documentation and receive relevant excerpts with source links. Use it to find information about workflows, features, and integration patterns.

The MCP server is best suited for discovering documentation and understanding concepts. For exact API specifications (methods, paths, parameters, required fields, enums, and schemas), use the OpenAPI contracts as the source of truth.

Documentation index ​

Treezor supports the llms.txt proposal. Start with /llms.txt to discover which pages are relevant to your task. Retrieve only what you need rather than loading the full documentation.

OpenAPI specifications ​

Use guides for workflows and business logic. Verify exact HTTP methods, paths, parameters, required fields, enums, and schemas against the OpenAPI contracts:

The API References section provides interactive Scalar viewers for exploring endpoints — these are for human use, not agent ingestion.

Bulk documentation file ​

/llms-full.txt contains the complete documentation in a single file (~1.9 MB, ~48K lines). Use this when your agent can search or parse local files more efficiently than making repeated HTTP requests. For interactive work, prefer the MCP server or the index plus targeted retrieval.

Suggested agent instruction ​

text
Use the Treezor MCP server to search documentation, or /llms.txt as the 
documentation index if MCP is not available.

Retrieve only the pages relevant to the task. Use Treezor guides for 
workflows and business behavior. Verify exact methods, paths, parameters, 
required fields, enums, and schemas against the OpenAPI specifications.

Do not invent undocumented fields or behavior. Report any conflict between 
a guide and an API contract.

Limitations ​

  • The MCP server searches indexed documentation content; it does not have access to OpenAPI specifications
  • The llms-full.txt file may contain VitePress formatting artifacts
  • Some internal links in the Markdown output may not resolve when fetched directly
  • OpenAPI files are renamed with date suffixes on deployment — the links above point to the source files

Updated on: