LLM Relay Doctor

One command tells you why your API / relay endpoint is broken, in 20 seconds.

Tests baseURL, key, model list, SSE streaming, first-token latency, 180s-hang class of bugs
Detects the exact failure mode: 308 loop, silent stall, credential mismatch, transport mismatch
Outputs a copy-pasteable config patch for the working path
Ships with a written troubleshooting playbook from real relay outages

29 USDT

Pay in USDT (BEP20). The download link unlocks the moment the transfer confirms — no account needed.

What you get

A single Python file plus a one-page README — no framework, no account, no license server.
Sample data and a runnable demo, so you can see the output before you read the code.
Source included. Runs offline on Python 3.9+. Most tools use the standard library only.

How to run it

unzip llm-relay-fixer.zip && cd llm-relay-fixer
pip install -r requirements.txt   # only if the README lists one
python llm-relay-fixer.py demo/input.csv -o out.csv

One-time payment, no subscription, no seat count. If the file will not run on your machine, reply to the order and it gets fixed or refunded.

Background reading

Your OpenAI-compatible endpoint hangs for 180 seconds, then fails: the four real causes