Self-hosted, OpenAI-compatible LLM gateway that measures the cost of every request and attributes it to the customer, workflow, or feature that drove it — without ever storing the prompt or response. Change one base_url; every call produces a payload-free economic receipt. Apache-2.0, zero dependency on any Inferrail-operated service.
Instructions file
Claude Code instructions for domondi1/inferrail, covering inferrail — agent context, before major feature work, read, before implementing something substantial, ask, engineering rules and repo-level conventions.