v1.1: Warme agent — directe Python import i.p.v. subprocess

- Vervang subprocess calls naar hermes chat -q door directe Python import
- Nieuwe hermes_client.py met warme AIAgent (eenmalig init, hergebruik per query)
- Initializing agent overhead: 3-5s per query -> 0.48s eenmalig
- Queries na eerste zijn ~5s (alleen LLM API tijd)
- Auxiliary vision provider leeggezet (geen DNS lookup naar Nous)

app.py

@@ -1,47 +1,53 @@
 """
-Hermes Relay — Flask bridge between Node-RED and Hermes Agent.
+Hermes Relay — Flask bridge tussen Node-RED en Hermes Agent (v2.0).
 
-Node-RED sends a POST with msg.payload → this script calls `hermes chat -q`
-via subprocess → returns the answer as JSON response AND sends it to the
-Telegram chat via `hermes send`.
+Gebruikt een warme Hermes agent (directe Python import) in plaats van
+subprocess per query. Dit bespaart ~3-5 seconden startup tijd per query.
+
+Node-RED stuurt POST met msg.payload → Hermes (direct) → Antwoord + Telegram.
 
 Endpoints:
   POST /ask       — main relay endpoint
-  GET  /health    — health check for Node-RED monitoring
-
-Architecture:
-  Node-RED (192.168.1.125) ──POST──▶ Hermes Relay (192.168.1.74:8650)
-                                       │
-                                       ├──▶ hermes chat -q (subprocess)
-                                       │         │
-                                       │         ▼
-                                       │    JSON response ──▶ Node-RED
-                                       │
-                                       └──▶ hermes send --to telegram
-                                                  │
-                                                  ▼
-                                             Telegram chat
+  GET  /health    — health check
 """
-
-import subprocess
+import os
 import sys
 import time
+import logging
+
+# ── Hermes path toevoegen (zodat we Hermes direct kunnen importeren) ───
+_HERMES_SITE = '/usr/local/lib/hermes-agent/venv/lib/python3.11/site-packages'
+if _HERMES_SITE not in sys.path:
+    sys.path.insert(0, _HERMES_SITE)
+
 from flask import Flask, request, jsonify
 
+# ── Config ──────────────────────────────────────────────────────────────
+TIMEOUT = 120          # max seconden voor Hermes query
+TG_TARGET = "telegram" # hermes send target
+PROMPT_PREFIX = "/fast" # wordt aan elke prompt toegevoegd
+MODEL = "qwen3.7-max"   # model override (leeg = default uit config)
+
+# ── Logging ─────────────────────────────────────────────────────────────
+logging.basicConfig(
+    level=logging.INFO,
+    format="[hermes-relay] %(asctime)s %(levelname)s %(message)s",
+    stream=sys.stderr,
+)
+logger = logging.getLogger(__name__)
+
+# ── Flask app ───────────────────────────────────────────────────────────
 app = Flask(__name__)
 
-# ── Config ──────────────────────────────────────────────────────────────
-HERMES_BIN = "/usr/local/bin/hermes"
-TIMEOUT = 120  # seconds — hermes can take a while on complex queries
-TG_TARGET = "telegram"  # hermes send target (home channel)
-
-# Prompt & Model
-PROMPT_PREFIX = "/fast"  # prepended to every query (empty string to disable)
-MODEL = "qwen3.7-max"    # model override (empty string to use default)
+# ── Hermes warm houden ──────────────────────────────────────────────────
+# Import de Hermes client (triggert eenmalige module laden)
+HERMES_BIN = "/usr/local/bin/hermes"  # nog steeds nodig voor hermes send
+from hermes_client import call_hermes as _hermes_query
 
 
 def send_telegram(text: str) -> bool:
-    """Send a message via hermes send. Returns True on success."""
+    """Stuur bericht via hermes send subprocess."""
+    import subprocess
     try:
         result = subprocess.run(
             [HERMES_BIN, "send", "--to", TG_TARGET, "--quiet", text],
@@ -51,45 +57,18 @@ def send_telegram(text: str) -> bool:
         )
         return result.returncode == 0
     except Exception as e:
-        print(f"[hermes-relay] Telegram send failed: {e}", file=sys.stderr)
+        logger.error(f"Telegram send failed: {e}")
         return False
 
 
-def call_hermes(prompt: str) -> str:
-    """Call hermes chat -q and return stdout."""
-    # Build full prompt with prefix
-    full_prompt = f"{PROMPT_PREFIX} {prompt}".strip() if PROMPT_PREFIX else prompt
-
-    cmd = [HERMES_BIN, "chat", "-q", full_prompt, "-Q", "--yolo"]
-
-    # Add model override if configured
-    if MODEL:
-        cmd.extend(["-m", MODEL])
-
-    try:
-        result = subprocess.run(
-            cmd,
-            capture_output=True,
-            text=True,
-            timeout=TIMEOUT,
-        )
-        if result.returncode != 0:
-            return f"⚠️ Hermes exited with code {result.returncode}\nstderr: {result.stderr.strip()}"
-        return result.stdout.strip()
-    except subprocess.TimeoutExpired:
-        return f"⚠️ Timeout — Hermes took longer than {TIMEOUT}s"
-    except Exception as e:
-        return f"⚠️ Error: {e}"
-
-
 # ── Routes ──────────────────────────────────────────────────────────────
 
 @app.route("/ask", methods=["POST"])
 def ask():
-    """Main relay endpoint. Expects JSON body with 'payload' field."""
+    """Main relay endpoint."""
     start = time.time()
 
-    # Accept both JSON body and form data
+    # Parse input
     if request.is_json:
         data = request.get_json(silent=True) or {}
         prompt = data.get("payload", "")
@@ -99,15 +78,15 @@ def ask():
     if not prompt:
         return jsonify({"status": "error", "message": "No payload received"}), 400
 
-    print(f"[hermes-relay] Received: {prompt[:100]}...")
+    logger.info(f"Received: {prompt[:100]}...")
 
-    # Call Hermes
-    answer = call_hermes(prompt)
+    # Roep Hermes aan (direct, geen subprocess)
+    answer = _hermes_query(prompt)
     elapsed = round(time.time() - start, 1)
 
-    print(f"[hermes-relay] Answer ({elapsed}s): {answer[:200]}...")
+    logger.info(f"Answer ({elapsed}s): {answer[:200]}...")
 
-    # Also send to Telegram
+    # Stuur ook naar Telegram
     tg_msg = f"🤖 Hermes Relay\n\nVraag: {prompt}\n\nAntwoord: {answer}"
     tg_sent = send_telegram(tg_msg)
 
@@ -121,16 +100,15 @@ def ask():
 
 @app.route("/health", methods=["GET"])
 def health():
-    """Health check — use this in Node-RED to monitor relay status."""
+    """Health check voor Node-RED monitoring."""
     return jsonify({
         "status": "ok",
         "service": "hermes-relay",
-        "hermes_bin": HERMES_BIN,
-        "timeout": TIMEOUT,
-        "prompt_prefix": PROMPT_PREFIX or "(none)",
         "model": MODEL or "(default)",
+        "prompt_prefix": PROMPT_PREFIX or "(none)",
     })
 
 
 if __name__ == "__main__":
+    logger.info("Starting Hermes Relay v2.0 (warme agent)...")
     app.run(host="0.0.0.0", port=8650, debug=False)

hermes_client.py

@@ -0,0 +1,131 @@
+"""
+Hermes Client — directe Python import (geen subprocess overhead).
+
+Warme agent: eenmalig initialiseren, daarna hergebruiken voor elke query.
+Dit bespaart de subprocess + Python startup overhead van 2-5 seconden per query.
+"""
+import io
+import os
+import sys
+import time
+import logging
+from contextlib import redirect_stdout, redirect_stderr
+
+logger = logging.getLogger(__name__)
+
+# ── Eenmalige Hermes imports (bij module laden) ──────────────────────
+_HERMES_LOADED = False
+_cli = None
+
+
+def _ensure_hermes_loaded():
+    """Laad Hermes eenmalig bij eerste gebruik — dit is het zware werk."""
+    global _HERMES_LOADED, _cli
+    if _HERMES_LOADED:
+        return
+
+    start = time.time()
+
+    # Zorg dat Hermes in de Python path zit
+    hermes_path = '/usr/local/lib/hermes-agent/venv/lib/python3.11/site-packages'
+    if hermes_path not in sys.path:
+        sys.path.insert(0, hermes_path)
+
+    # Force UTF-8
+    os.environ.setdefault('PYTHONIOENCODING', 'utf-8')
+
+    # Laad .env via Hermes' eigen loader
+    from hermes_cli.env_loader import load_hermes_dotenv
+    from hermes_constants import get_hermes_home
+    _hermes_home = get_hermes_home()
+    _project_env = os.path.join(os.path.dirname(hermes_path), '.env')
+    load_hermes_dotenv(hermes_home=_hermes_home, project_env=_project_env)
+
+    # Load config
+    from hermes_cli.config import load_config
+    config = load_config()
+
+    # Resolve toolsets (zelfde als 'cli' platform)
+    from hermes_cli.tools_config import _get_platform_tools
+    toolsets = sorted(_get_platform_tools(config, 'cli'))
+
+    # YOLO mode + interactive mode
+    os.environ['HERMES_YOLO_MODE'] = '1'
+    os.environ['HERMES_INTERACTIVE'] = '1'
+    os.environ['HERMES_SESSION_SOURCE'] = 'relay'
+
+    # Import HermesCLI
+    from cli import HermesCLI
+
+    # Maak CLI instantie (nog géén agent init — dat gebeurt pas bij eerste query)
+    _cli = HermesCLI(
+        model=config.get('model', {}).get('default', ''),
+        toolsets=toolsets,
+    )
+    # Onderdruk banner/status output
+    _cli.tool_progress_mode = 'off'
+
+    elapsed = time.time() - start
+    logger.info(
+        f"Hermes loaded in {elapsed:.2f}s "
+        f"(toolsets: {toolsets})"
+    )
+    _HERMES_LOADED = True
+
+
+def call_hermes(prompt: str) -> str:
+    """
+    Roep Hermes aan — hergebruikt een warme agent.
+
+    De agent wordt één keer geïnitialiseerd (bij eerste query) en daarna
+    hergebruikt. Na elke query wordt de conversation history gereset,
+    zodat elke query een frisse start krijgt.
+    """
+    _ensure_hermes_loaded()
+
+    # Prefix (zoals /fast)
+    prefix = os.environ.get('HERMES_RELAY_PREFIX', '/fast')
+    full_prompt = f"{prefix} {prompt}".strip() if prefix else prompt
+
+    start = time.time()
+
+    # ── Agent initialiseren (eenmalig, cached) ──
+    turn_route = _cli._resolve_turn_agent_config(full_prompt)
+    if turn_route['signature'] != _cli._active_agent_route_signature:
+        _cli.agent = None
+    init_ok = _cli._init_agent(
+        model_override=turn_route['model'],
+        runtime_override=turn_route['runtime'],
+        request_overrides=turn_route.get('request_overrides'),
+    )
+    if not init_ok:
+        return "⚠️ Kon Hermes agent niet initialiseren."
+
+    # Stil modus
+    _cli.agent.quiet_mode = True
+    _cli.agent.suppress_status_output = True
+    _cli.agent.stream_delta_callback = None
+    _cli.agent.tool_gen_callback = None
+
+    # ── Query uitvoeren ──
+    # Capture alles wat naar stdout/stderr gaat (session_id, banner, etc.)
+    stdout_capture = io.StringIO()
+    stderr_capture = io.StringIO()
+
+    try:
+        with redirect_stdout(stdout_capture), redirect_stderr(stderr_capture):
+            # conversation_history wissen voor frisse start
+            _cli.agent.messages = []
+            result = _cli.agent.chat(full_prompt)
+    except SystemExit:
+        # Hermes roept sys.exit() aan in sommige error paths
+        logger.warning("Hermes riep sys.exit() aan tijdens query")
+        result = stdout_capture.getvalue().strip()
+    except Exception as e:
+        logger.error(f"Hermes error: {e}", exc_info=True)
+        return f"⚠️ Hermes error: {e}"
+
+    elapsed = time.time() - start
+    logger.info(f"Hermes responded in {elapsed:.1f}s ({len(result)} chars)")
+
+    return result