refactor(reasoning)!: remove Node-RED engine and integrations

BREAKING CHANGE: NODE_RED reasoning is no longer supported. Unsupported
mode values fall back to PROXY. PROXY and OPENAI remain available.

s2s-client/s2s_client/core/config.py

@@ -15,8 +15,6 @@ def _normalize_reasoning_type(value: str) -> str:
     normalized = (value or "").strip().strip('"').strip("'").replace("-", "_").upper()
     if normalized == "OPENAI":
         return "OPENAI"
-    if normalized in {"NODE_RED", "NODERED"}:
-        return "NODE_RED"
     return "PROXY"
 
 
@@ -124,8 +122,7 @@ if S2S_REPORTS_INJECT_MODE not in {"minimal", "full", "off", "none"}:
 # Feature flag: when "true" the proxy logs the FULL PROMPT TEXT (not just
 # lengths) going to the upstream LLM/realtime backends. Includes:
 #   - WS session.update session.instructions (before/after patch)
-#   - /api/reasoning NODE_RED instructions + user_context full text
-#   - /api/reasonING OPENAI full system message + user transcript
+#   - /api/reasoning OPENAI full system message + user transcript
 # Default = off; only length summaries are logged (current behaviour).
 # WARNING: setting this on will log potentially PII / sensitive caller
 # profile data / user transcripts. Use only in development / debugging.
@@ -186,9 +183,6 @@ APIPILOT_HTTP_TIMEOUT = max(1.0, min(120.0, _APIPILOT_TO_RAW))
 del _APIPILOT_TO_RAW
 
 
-# Node-RED reasoning engine endpoint (NODE_RED mode).
-# Examples: http://localhost:1881 or http://s2s-reasoning:1880 (Docker network).
-NODE_RED_REASONING_URL = os.environ.get("S2S_REASONING_NODE_RED_URL", "").strip().rstrip("/")
 # Speech-to-speech load balancer URL. When set, the browser POSTs /api/session
 # (which proxies <lb>/session here, server-side) and connects to the URL the LB
 # returns (the original flow). The LB address itself is never sent to the browser.