API reference

Auto-generated

This page is auto-generated from the gateway's OpenAPI specification. It is rebuilt on every gateway release by the openapi-sync workflow.

The page below is a placeholder until the first gateway release tag triggers the sync. After that, this content is overwritten with the rendered widdershins output.

Endpoints (placeholder structure)

POST /v1/chat/completions

Create a chat completion with a per-query environmental receipt.

Authentication: Bearer token (see Authentication)

Request body:

Field	Type	Required	Description
model	string	yes	Model identifier (see /v1/models)
messages	array	yes	Chat messages
temperature	number	no	0.0–2.0; default 1.0
top_p	number	no	Nucleus sampling
max_tokens	integer	no	Max completion tokens
stream	boolean	no	Stream the response
extra_body.vetted.tier	string	no	auto (default), proxy, parametric, telemetry
extra_body.vetted.boundary	string	no	comprehensive (default) or narrow
extra_body.vetted.region	string	no	Pin to a specific region

Response: OpenAI Chat Completion shape with additional vetted block. See first call.

POST /v1/embeddings

Create text embeddings with a per-call environmental receipt.

[Full schemas, request/response examples, and error tables — populated by widdershins on next sync.]