For AI Agents & LLMs

Machine-readable specs and context files for integrating RaxyProxy into Claude, Cursor, ChatGPT, and other AI tools.

Claude

ChatGPT

Cursor

Gemini

LangChain

Llama

RaxyProxy publishes a Claude Code skill and machine-readable documentation for use with Claude, Cursor, ChatGPT, and automated tooling.

Claude Code Skill

Install the raxyproxy skill into Claude Code, Cursor, Gemini CLI, or any agent that supports the Agent Skills standard:

npx skills add raxyproxy/skills

Then set your API key (get it from API Keys):

export RAXYPROXY_API_KEY="raxy_..."

The skill teaches your agent how to generate proxy strings, list your packages, apply geo-targeting, and use sticky sessions — all by calling the RaxyProxy API with your key. Invoke it with /raxyproxy or let the agent detect it automatically from your prompt.

Machine-readable specs

llms.txt

https://raxyproxy.com/llms.txt

A plain-text context document following the llms.txt web convention (analogous to robots.txt — the name is fixed by the standard). Paste it as a Claude Project instruction or IDE workspace context to give the model complete knowledge of RaxyProxy's proxy types, pricing, geo-targeting syntax, and connection format.

llms-api.md / llms-api.txt

https://raxyproxy.com/llms-api.md

A Markdown-formatted reference covering every REST API endpoint — parameters, response shapes, proxy format strings, error codes, and geo-targeting rules. Optimised for token efficiency. Use it as a Claude Project document, a Cursor rule, or paste it directly into a conversation.

# RaxyProxy API reference — fetched from llms-api.md
@https://raxyproxy.com/llms-api.md

openapi.json

https://raxyproxy.com/openapi.json

Full OpenAPI 3.0 specification. Import it into:

• Postman / Insomnia — import the URL directly to generate a collection
• Swagger UI — paste the URL to get interactive docs
• openapi-generator-cli — generate typed clients in any language
• Claude / ChatGPT custom actions — reference as the API schema

Using with Claude & Cursor

# Download for local use
curl -o llms-api.md  https://raxyproxy.com/llms-api.md
curl -o openapi.json https://raxyproxy.com/openapi.json

For AI agent pipelines that need proxy credentials at runtime, fetch /llms.txt once at startup — it contains the full connection format, geo-targeting syntax, and billing rules so the agent can construct proxy strings without additional lookups. Credentials are available from your API keys page.