Module llm_local 

Optional one-shot LLM HTTP extract (XDG key only). One-shot optional LLM HTTP extract (XDG key only; no product env vars).

Uses an OpenAI-compatible chat completions endpoint configured via XDG (openrouter_api_key, llm_base_url, llm_model). No telemetry.

Constants

DEFAULT_LLM_BASE_URL

Default OpenAI-compatible base URL (path ends before /chat/completions).

DEFAULT_LLM_MODEL

Default model id when XDG llm_model is unset.

Functions

base_url

Base URL from XDG or default constant.

chat_completion

Call chat completions with retry/backoff (one-shot; no daemon).

extract_with_llm

Build extract+LLM payload from free text and optional question/schema.

model

Model from XDG or default constant.

require_api_key

Resolve API key from XDG only.