Technical Spec v1.0

Gateway Specification

The Lindr Gateway is a high-performance, protocol-aware proxy that instruments LLM traffic at the edge.

Network & Protocol

Base URL

https://gateway.lindr.io/v1

Protocol

HTTPS / TLS 1.3

Authentication Headers

Header	Description	Required
`X-Lindr-API-Key`	Your Lindr organization API key	Yes
`X-Lindr-Monitor-Id`	The ID of the target monitor	Yes
`X-Lindr-Project-Id`	Optional project scope	No

Provider Compatibility

The gateway is fully compatible with the OpenAI Chat Completions API. Requests following this spec can be routed to OpenAI, Anthropic, or Google via Lindr without schema changes.

OpenAI

Anthropic

Google Gemini

Azure OpenAI

Llama (Self-hosted)

Payload Specification

The gateway expects a standard JSON body. The messages array is used to extract context for lexical and semantic analysis.

{
  "model": "gpt-4o",
  "messages": [
    { "role": "system", "content": "You are a helpful assistant." },
    { "role": "user", "content": "Hello!" }
  ],
  "temperature": 0.7,
  "stream": false
}