For AI agents
Need answers about your company’s data rather than code? See Connect to Claude and ChatGPT.
The portal serves the documentation in formats convenient for language models and coding agents.
| What | Address |
|---|---|
| A prompt that sets an agent up to work with the API: project rules and a copy of the specification | /en/agent-setup/prompt.md; description in AI agent setup |
| A documentation index in the llms.txt standard (in Russian) | /llms.txt |
| All documentation in one file (in Russian) | /llms-full.txt |
| A markdown version of any page | the page address + .md, for example /en/start/first-request.md |
| The OpenAPI specification: exact parameters, schemas and response codes of all methods | /en/openapi.yaml |
The API reference is not included in llms.txt or in the markdown versions: it is generated from the specification,
and for agents the source of truth on methods is the specification itself, /openapi.yaml.
The “Copy page” menu
Section titled “The “Copy page” menu”Each page title has a menu: copy the page as markdown, open the markdown in a new tab, save as PDF via the browser’s print dialog, open the page in ChatGPT, Claude, Gemini, Cursor and other assistants.
How to connect an agent
Section titled “How to connect an agent”The shortest path is the setup prompt: the agent downloads the specification itself and writes the rules
into the project rules file. Alternatively, give the agent a link to /llms.txt: it is an index of pages (in Russian) with links to their markdown versions.
To generate code for a specific method, pass /openapi.yaml or a fragment of it:
the descriptions in the specification are complete, while those on the reference pages are simplified.