Add tracked nginx reverse-proxy template for llama-server — previously only existed as a heredoc run live on BigBoy; also resolves the server_name wildcard conflict warning

ansible/templates/llama.conf.j2

@@ -0,0 +1,39 @@
+# llama.conf — nginx reverse proxy for BigBoy's llama-server
+#
+# This is a Jinja2 template (Ansible's `template` module convention),
+# ready to use once Phase 5 is formalized into an actual Ansible role.
+# Until then, it's the tracked, reproducible reference for what's
+# actually configured live on BigBoy — this previously existed only as
+# a heredoc typed directly during the deployment session, never
+# committed anywhere.
+#
+# Manual deployment (until this is wired into a real Ansible role):
+#   1. Fill in the Jinja2 variables below with their actual values
+#   2. Copy the rendered result to /etc/nginx/conf.d/llama.conf
+#   3. sudo nginx -t   (confirm syntax before reloading)
+#   4. sudo systemctl reload nginx
+#
+# NOTE on server_name: the first live deployment used `server_name _;`
+# (the catch-all wildcard) and hit a harmless-but-noisy warning —
+# "conflicting server name "_" on 0.0.0.0:80, ignored" — because
+# AlmaLinux's stock nginx package already ships its own default catch-all
+# server block using the same wildcard. This template uses the actual
+# host IP instead, avoiding the duplicate entirely rather than just
+# tolerating the warning.
+
+server {
+    listen 80;
+    server_name {{ ip_address }};
+
+    location / {
+        proxy_pass http://{{ llama_server_host }}:{{ llama_server_port }};
+
+        # Required for llama-server's streaming (server-sent event)
+        # responses — without these, chat responses arrive all at once
+        # instead of streaming token-by-token.
+        proxy_http_version 1.1;
+        proxy_set_header Connection "";
+        proxy_buffering off;
+        proxy_read_timeout 300s;
+    }
+}