fix(litellm): route kimi-k3 to the Kimi Code (coding) endpoint

The Heid panel plan uses Kimi's coding endpoint, not the general Moonshot API.
kimi-k3 now → openai/k3 @ https://api.kimi.com/coding/v1 (KIMI_CODE_API_KEY,
Vivace); the original general-endpoint entry is kept as kimi-k3-gen-api
(api.moonshot.ai, MOONSHOT_API_KEY). Both verified live through the gateway.
Same k3 constraints on both: temperature MUST be 1 (else 400), reasoning model
(reasoning_content vs content, needs adequate max_tokens).

stacks/litellm/.env.example

@@ -35,7 +35,12 @@ VLLM_API_KEY=
 # (glm-5.1 / glm-5-turbo / glm-4.7 / glm-4.5-air).
 Z_AI_API_KEY=
 
-# Moonshot AI (Kimi) key — fronts kimi-k3 (https://api.moonshot.ai/v1). PAID.
+# Kimi (Moonshot) keys — PAID.
+#   KIMI_CODE_API_KEY — the CODING endpoint (https://api.kimi.com/coding/v1),
+#     Vivace membership; fronts the primary `kimi-k3` arm (upstream model `k3`).
+#   MOONSHOT_API_KEY  — the general endpoint (https://api.moonshot.ai/v1);
+#     fronts the `kimi-k3-gen-api` variant.
+KIMI_CODE_API_KEY=
 MOONSHOT_API_KEY=
 
 # --- Langfuse-ready (leave blank for the lean first cut) ---