Method Chat

internal/llm/llm.go:286–290

Chat streams an assistant response on the returned channel, closing it when the stream ends. Reasoning runs at `high` effort by default; if the server rejects the tools + reasoning_effort combo (newer OpenAI models do), postChat drops reasoning_effort for this Client's lifetime so the model still wo

(parent context.Context, messages []chmctx.Message, tools []Tool)

// drops reasoning_effort for this Client's lifetime so the model still works,
// with tools but no reasoning. Staying on chat-completions is the product line;
// we do not branch to /v1/responses to keep reasoning.
func (c *Client) Chat(parent context.Context, messages []chmctx.Message, tools []Tool) <-chan Event {
	out := make(chan Event, 32)
	go c.run(parent, messages, tools, out)
	return out
}

func (c *Client) run(parent context.Context, msgs []chmctx.Message, tools []Tool, out chan<- Event) {