Chat Completions Openai Compatible

Operator reuses an existing OpenAI SDK client and forgets to override base_url, so chat completion requests with a Groq API key go to api.openai.com instead of GroqCloud.

Point the OpenAI-compatible client at base_url https://api.groq.com/openai/v1 and pass the GROQ_API_KEY as the Bearer key. Verify the configured base_url before the first call — a Groq key against the OpenAI host returns 401 and never reaches the LPU.

Agent builds messages[] as system, assistant, user (assistant turn before any user turn) when constructing the first request.

Build messages[] with a leading system message (optional) followed by a user turn; alternate user/assistant thereafter. GroqCloud follows the OpenAI Chat Completions contract — the conversation must open with system and/or user content, not a bare assistant turn.

Agent sets max_tokens=256 for a detailed code-generation prompt. The response comes back with finish_reason=length and an obviously truncated function body.

Detect choices[0].finish_reason=length and either surface the output as partial or issue a continuation request (append the truncated assistant turn plus a 'continue' instruction). Never present a length-truncated completion as a complete answer.