fix: default LLM_API_BASE to host.docker.internal for local Ollama

Most developers run Ollama locally via `ollama serve` on the host machine
with Kind/Docker Desktop. The previous default (ollama.ollama.svc:11434)
assumed Ollama was deployed as a Kubernetes service, causing connection
failures for the common local setup.

Change the default to host.docker.internal:11434 and document the
in-cluster alternative as a comment.

Signed-off-by: Mariusz Sabath <mrsabath@gmail.com>
Co-authored-by: Cursor <cursoragent@cursor.com>

Author: mrsabath

a2a/git_issue_agent/.env.ollama

@@ -6,7 +6,9 @@
 
 # LLM configuration
 TASK_MODEL_ID=ollama/ibm/granite4:latest
-LLM_API_BASE=http://ollama.ollama.svc:11434
+# Default: local Ollama via Kind/Docker Desktop host gateway.
+# For in-cluster Ollama: LLM_API_BASE=http://ollama.ollama.svc:11434
+LLM_API_BASE=http://host.docker.internal:11434
 LLM_API_KEY=ollama
 MODEL_TEMPERATURE=0