"""
Bot Gateway HTTP API
Handles incoming webhooks from Telegram, Discord, etc.
"""
import asyncio
import base64
import json
import re
import logging
import os
import sys
import time
import uuid
import httpx
from pathlib import Path
from typing import Dict, Any, Optional, List, Tuple
from datetime import datetime
from dataclasses import dataclass
from io import BytesIO

from fastapi import APIRouter, HTTPException
from pydantic import BaseModel

from router_client import send_to_router
from memory_client import memory_client
from services.doc_service import (
    parse_document,
    ingest_document,
    ask_about_document,
    get_doc_context
)
from behavior_policy import (
    should_respond,
    analyze_message,
    detect_media_question,
    detect_explicit_request,
    detect_url,
    detect_agent_mention,
    is_no_output_response,
    record_interaction,
    record_ack,
    get_ack_text,
    is_prober_request,
    has_agent_chat_participation,
    NO_OUTPUT,
    BehaviorDecision,
    AGENT_NAME_VARIANTS,
)

logger = logging.getLogger(__name__)

# Telegram message length limits
TELEGRAM_MAX_MESSAGE_LENGTH = 4096
TELEGRAM_SAFE_LENGTH = 3500  # Leave room for formatting
# Operator pending state cache (chat_id -> {ts, items})
LAST_PENDING_STATE: Dict[str, Dict[str, Any]] = {}
PENDING_STATE_TTL = 1800  # 30 minutes


# Per-user language preference cache (chat_id:user_id -> {lang, ts})
USER_LANGUAGE_PREFS: Dict[str, Dict[str, Any]] = {}
USER_LANGUAGE_PREF_TTL = 30 * 24 * 3600  # 30 days

# Recent photo context for follow-up questions in chat (agent:chat:user -> {file_id, ts})
RECENT_PHOTO_CONTEXT: Dict[str, Dict[str, Any]] = {}
RECENT_PHOTO_TTL = 30 * 60  # 30 minutes


def _cleanup_recent_photo_context() -> None:
    now = time.time()
    expired = [k for k, v in RECENT_PHOTO_CONTEXT.items() if now - float(v.get("ts", 0)) > RECENT_PHOTO_TTL]
    for k in expired:
        del RECENT_PHOTO_CONTEXT[k]


def _set_recent_photo_context(agent_id: str, chat_id: str, user_id: str, file_id: str) -> None:
    _cleanup_recent_photo_context()
    key = f"{agent_id}:{chat_id}:{user_id}"
    RECENT_PHOTO_CONTEXT[key] = {"file_id": file_id, "ts": time.time()}


def _get_recent_photo_file_id(agent_id: str, chat_id: str, user_id: str) -> Optional[str]:
    _cleanup_recent_photo_context()
    key = f"{agent_id}:{chat_id}:{user_id}"
    rec = RECENT_PHOTO_CONTEXT.get(key)
    if not rec:
        return None
    return rec.get("file_id")


def _looks_like_photo_followup(text: str) -> bool:
    if not text:
        return False
    t = text.strip().lower()
    markers = [
        "що ти бачиш", "що на фото", "що на зображенні", "опиши фото", "подивись фото",
        "what do you see", "what is in the image", "describe the photo",
        "что ты видишь", "что на фото", "опиши фото", "посмотри фото",
    ]
    return any(m in t for m in markers)


def _cleanup_user_language_prefs() -> None:
    now = time.time()
    expired = [k for k, v in USER_LANGUAGE_PREFS.items() if now - float(v.get("ts", 0)) > USER_LANGUAGE_PREF_TTL]
    for k in expired:
        del USER_LANGUAGE_PREFS[k]


def _normalize_lang_code(raw: Optional[str]) -> Optional[str]:
    if not raw:
        return None
    code = str(raw).strip().lower().replace("_", "-")
    if code.startswith("uk"):
        return "uk"
    if code.startswith("ru"):
        return "ru"
    if code.startswith("en"):
        return "en"
    return None


def _detect_language_from_text(text: str) -> Optional[str]:
    if not text:
        return None
    t = text.lower()
    letters = [ch for ch in t if ch.isalpha()]
    if not letters:
        return None

    cyr = sum(1 for ch in letters if "а" <= ch <= "я" or ch in "іїєґё")
    lat = sum(1 for ch in letters if "a" <= ch <= "z")

    if cyr >= 3 and cyr >= lat:
        # Ukrainian-specific letters strongly indicate Ukrainian.
        if any(ch in t for ch in "іїєґ"):
            return "uk"
        # Russian-specific letters/symbols.
        if any(ch in t for ch in "ёыэъ"):
            return "ru"
        # Soft lexical preference.
        uk_hits = sum(1 for w in ("що", "який", "дякую", "будь", "будь ласка", "привіт") if w in t)
        ru_hits = sum(1 for w in ("что", "какой", "спасибо", "пожалуйста", "привет") if w in t)
        if uk_hits > ru_hits:
            return "uk"
        if ru_hits > uk_hits:
            return "ru"
        return "uk"

    if lat >= 3 and lat > cyr:
        return "en"

    return None


def resolve_preferred_language(chat_id: str, user_id: str, text: str, telegram_lang_code: Optional[str]) -> str:
    _cleanup_user_language_prefs()
    key = f"{chat_id}:{user_id}"

    text_lang = _detect_language_from_text(text)
    tg_lang = _normalize_lang_code(telegram_lang_code)
    cached_lang = USER_LANGUAGE_PREFS.get(key, {}).get("lang")

    preferred = text_lang or tg_lang or cached_lang or "uk"
    USER_LANGUAGE_PREFS[key] = {"lang": preferred, "ts": time.time()}
    return preferred


def preferred_language_label(lang: str) -> str:
    return {
        "uk": "Ukrainian",
        "ru": "Russian",
        "en": "English",
    }.get((lang or "").lower(), "Ukrainian")


def _extract_preferred_language_from_profile_fact(fact: Optional[Dict[str, Any]]) -> Optional[str]:
    if not isinstance(fact, dict):
        return None
    data = fact.get("fact_value_json")
    if not isinstance(data, dict):
        return None
    preferred = _normalize_lang_code(data.get("preferred_language"))
    if preferred:
        return preferred
    return _normalize_lang_code(data.get("language_code"))


async def resolve_preferred_language_persistent(
    chat_id: str,
    user_id: str,
    text: str,
    telegram_lang_code: Optional[str],
    team_id: Optional[str] = None,
) -> str:
    """Resolve language with memory-service fallback for post-restart continuity."""
    _cleanup_user_language_prefs()
    key = f"{chat_id}:{user_id}"

    text_lang = _detect_language_from_text(text)
    tg_lang = _normalize_lang_code(telegram_lang_code)
    cached_lang = USER_LANGUAGE_PREFS.get(key, {}).get("lang")

    if text_lang or tg_lang or cached_lang:
        preferred = text_lang or tg_lang or cached_lang or "uk"
        USER_LANGUAGE_PREFS[key] = {"lang": preferred, "ts": time.time()}
        return preferred

    try:
        fact = await memory_client.get_fact(
            user_id=f"tg:{user_id}",
            fact_key="profile",
            team_id=team_id,
        )
        fact_lang = _extract_preferred_language_from_profile_fact(fact)
        if fact_lang:
            USER_LANGUAGE_PREFS[key] = {"lang": fact_lang, "ts": time.time()}
            return fact_lang
    except Exception as e:
        logger.debug(f"preferred language fact lookup failed: {e}")

    USER_LANGUAGE_PREFS[key] = {"lang": "uk", "ts": time.time()}
    return "uk"


def _pending_state_cleanup():
    now = time.time()
    expired = [cid for cid, rec in LAST_PENDING_STATE.items() if now - rec.get('ts', 0) > PENDING_STATE_TTL]
    for cid in expired:
        del LAST_PENDING_STATE[cid]


def _get_last_pending(chat_id: str) -> list | None:
    _pending_state_cleanup()
    rec = LAST_PENDING_STATE.get(str(chat_id))
    if not rec:
        return None
    return rec.get('items')


def _set_last_pending(chat_id: str, items: list):
    LAST_PENDING_STATE[str(chat_id)] = {"ts": time.time(), "items": items}



def _chunk_text(text: str, max_len: int = 4096):
    if not text:
        return [""]
    chunks = []
    current = []
    current_len = 0
    for line in text.split("\n"):
        add_len = len(line) + (1 if current else 0)
        if current_len + add_len <= max_len:
            current.append(line)
            current_len += add_len
            continue
        if current:
            chunks.append("\n".join(current))
            current = []
            current_len = 0
        while len(line) > max_len:
            chunks.append(line[:max_len])
            line = line[max_len:]
        current.append(line)
        current_len = len(line)
    if current:
        chunks.append("\n".join(current))
    return chunks


# Training groups - agents respond to ALL messages without mention requirement
TRAINING_GROUP_IDS = {
    "-1003556680911",  # Agent Preschool Daarion.city
}

# Brand stack services
BRAND_INTAKE_URL = os.getenv("BRAND_INTAKE_URL", "http://brand-intake:9211").rstrip("/")
BRAND_REGISTRY_URL = os.getenv("BRAND_REGISTRY_URL", "http://brand-registry:9210").rstrip("/")
PRESENTATION_RENDERER_URL = os.getenv("PRESENTATION_RENDERER_URL", "http://presentation-renderer:9212").rstrip("/")
ARTIFACT_REGISTRY_URL = os.getenv("ARTIFACT_REGISTRY_URL", "http://artifact-registry:9220").rstrip("/")

router = APIRouter()


# ========================================
# Agent Configuration
# ========================================

@dataclass
class AgentConfig:
    """Конфігурація агента для стандартизації обробки повідомлень"""
    agent_id: str
    name: str
    prompt_path: str
    telegram_token_env: str
    default_prompt: str
    system_prompt: str = ""  # Буде встановлено після завантаження
    
    def load_prompt(self) -> str:
        """Завантажити system prompt з файлу"""
        try:
            p = Path(self.prompt_path)
            if not p.exists():
                logger.warning(f"{self.name} prompt file not found: {self.prompt_path}")
                return self.default_prompt
            
            prompt = p.read_text(encoding="utf-8")
            logger.info(f"{self.name} system prompt loaded ({len(prompt)} chars)")
            return prompt
        except Exception as e:
            logger.error(f"Error loading {self.name} prompt: {e}")
            return self.default_prompt
    
    def get_telegram_token(self) -> Optional[str]:
        """Отримати Telegram токен агента"""
        return os.getenv(self.telegram_token_env)


def load_agent_config(agent_id: str, name: str, prompt_path: str, 
                      telegram_token_env: str, default_prompt: str) -> AgentConfig:
    """Створити та завантажити конфігурацію агента"""
    config = AgentConfig(
        agent_id=agent_id,
        name=name,
        prompt_path=prompt_path,
        telegram_token_env=telegram_token_env,
        default_prompt=default_prompt,
        system_prompt=""  # Тимчасове значення
    )
    # Завантажити prompt
    config.system_prompt = config.load_prompt()
    return config


# ========================================
# Agent Configurations
# ========================================

# DAARWIZZ Configuration
DAARWIZZ_CONFIG = load_agent_config(
    agent_id="daarwizz",
    name=os.getenv("DAARWIZZ_NAME", "DAARWIZZ"),
    prompt_path=os.getenv(
        "DAARWIZZ_PROMPT_PATH",
        str(Path(__file__).parent / "daarwizz_prompt.txt"),
    ),
    telegram_token_env="DAARWIZZ_TELEGRAM_BOT_TOKEN",
    default_prompt=f"Ти — {os.getenv('DAARWIZZ_NAME', 'DAARWIZZ')}, AI-агент екосистеми DAARION.city. Допомагай учасникам з DAO-процесами."
)

# HELION Configuration
HELION_CONFIG = load_agent_config(
    agent_id="helion",
    name=os.getenv("HELION_NAME", "Helion"),
    prompt_path=os.getenv(
        "HELION_PROMPT_PATH",
        str(Path(__file__).parent / "helion_prompt.txt"),
    ),
    telegram_token_env="HELION_TELEGRAM_BOT_TOKEN",
    default_prompt=f"Ти — {os.getenv('HELION_NAME', 'Helion')}, AI-агент платформи Energy Union. Допомагай учасникам з технологіями та токеномікою."
)

# GREENFOOD Configuration
GREENFOOD_CONFIG = load_agent_config(
    agent_id="greenfood",
    name=os.getenv("GREENFOOD_NAME", "GREENFOOD"),
    prompt_path=os.getenv(
        "GREENFOOD_PROMPT_PATH",
        str(Path(__file__).parent / "greenfood_prompt.txt"),
    ),
    telegram_token_env="GREENFOOD_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — GREENFOOD Assistant, AI-ERP для крафтових виробників та кооперативів. Допомагай з обліком партій, логістикою, бухгалтерією та продажами."
)

# AGROMATRIX Configuration
AGROMATRIX_CONFIG = load_agent_config(
    agent_id="agromatrix",
    name=os.getenv("AGROMATRIX_NAME", "AgroMatrix"),
    prompt_path=os.getenv(
        "AGROMATRIX_PROMPT_PATH",
        str(Path(__file__).parent / "agromatrix_prompt.txt"),
    ),
    telegram_token_env="AGROMATRIX_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — AgroMatrix, AI-агент для агроаналітики, планування сезонів та кооперації фермерів. Допомагай з порадами щодо полів, процесів і ринків."
)

# ALATEYA Configuration
ALATEYA_CONFIG = load_agent_config(
    agent_id="alateya",
    name=os.getenv("ALATEYA_NAME", "Alateya"),
    prompt_path=os.getenv(
        "ALATEYA_PROMPT_PATH",
        str(Path(__file__).parent / "alateya_prompt.txt"),
    ),
    telegram_token_env="ALATEYA_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — Alateya, AI-агент R&D та біотех-інновацій. Допомагай з дослідженнями, протоколами й експериментальними дизайнами."
)

# NUTRA Configuration
NUTRA_CONFIG = load_agent_config(
    agent_id="nutra",
    name=os.getenv("NUTRA_NAME", "NUTRA"),
    prompt_path=os.getenv(
        "NUTRA_PROMPT_PATH",
        str(Path(__file__).parent / "nutra_prompt.txt"),
    ),
    telegram_token_env="NUTRA_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — NUTRA, нутріцевтичний агент платформи DAARION. Допомагаєш з формулами нутрієнтів, біомедичних добавок та лабораторних інтерпретацій. Консультуєш з питань харчування, вітамінів та оптимізації здоров'я."
)

# Registry of all agents (для легкого додавання нових агентів)
# 
# Щоб додати нового агента:
# 1. Створіть конфігурацію через load_agent_config():
#    NEW_AGENT_CONFIG = load_agent_config(
#        agent_id="new_agent",
#        name=os.getenv("NEW_AGENT_NAME", "New Agent"),
#        prompt_path=os.getenv("NEW_AGENT_PROMPT_PATH", str(Path(__file__).parent / "new_agent_prompt.txt")),
#        telegram_token_env="NEW_AGENT_TELEGRAM_BOT_TOKEN",
#        default_prompt="Ти — New Agent, AI-агент..."
#    )
# 2. Додайте до реєстру:
# DRUID Configuration
DRUID_CONFIG = load_agent_config(
    agent_id="druid",
    name=os.getenv("DRUID_NAME", "DRUID"),
    prompt_path=os.getenv(
        "DRUID_PROMPT_PATH",
        str(Path(__file__).parent / "druid_prompt.txt"),
    ),
    telegram_token_env="DRUID_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — DRUID, агент платформи DAARION. Допомагай користувачам з аналізом даних, рекомендаціями та інтеграцією RAG.",
)

# CLAN (Spirit) Configuration
CLAN_CONFIG = load_agent_config(
    agent_id="clan",
    name=os.getenv("CLAN_NAME", "Spirit"),
    prompt_path=os.getenv(
        "CLAN_PROMPT_PATH",
        str(Path(__file__).parent / "clan_prompt.txt"),
    ),
    telegram_token_env="CLAN_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — CLAN (Spirit), Дух Общини в екосистемі DAARION.city. Підтримуєш зв'язки між учасниками спільноти, зберігаєш традиції та допомагаєш в прийнятті колективних рішень.",
)

# EONARCH Configuration
EONARCH_CONFIG = load_agent_config(
    agent_id="eonarch",
    name=os.getenv("EONARCH_NAME", "EONARCH"),
    prompt_path=os.getenv(
        "EONARCH_PROMPT_PATH",
        str(Path(__file__).parent / "eonarch_prompt.txt"),
    ),
    telegram_token_env="EONARCH_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — EONARCH, провідник еволюції свідомості в екосистемі DAARION.city. Супроводжуєш людство на шляху трансформації свідомості до колективної мудрості.",
)

# SENPAI (Gordon Senpai) Configuration
SENPAI_CONFIG = load_agent_config(
    agent_id="senpai",
    name=os.getenv("SENPAI_NAME", "SENPAI"),
    prompt_path=os.getenv(
        "SENPAI_PROMPT_PATH",
        str(Path(__file__).parent / "senpai_prompt.txt"),
    ),
    telegram_token_env="SENPAI_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — Гордон Сенпай (Gordon Senpai), радник з ринків капіталу та цифрових активів. Допомагаєш з трейдингом, ризик-менеджментом, аналізом ринків.",
)

# 1OK Configuration
ONEOK_CONFIG = load_agent_config(
    agent_id="oneok",
    name=os.getenv("ONEOK_NAME", "1OK"),
    prompt_path=os.getenv(
        "ONEOK_PROMPT_PATH",
        str(Path(__file__).parent / "oneok_prompt.txt"),
    ),
    telegram_token_env="ONEOK_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — 1OK, асистент віконного майстра. Допомагаєш з кваліфікацією ліда, підготовкою заміру та формуванням комерційної пропозиції.",
)

# SOUL / Athena Configuration
SOUL_CONFIG = load_agent_config(
    agent_id="soul",
    name=os.getenv("SOUL_NAME", "Athena"),
    prompt_path=os.getenv(
        "SOUL_PROMPT_PATH",
        str(Path(__file__).parent / "soul_prompt.txt"),
    ),
    telegram_token_env="SOUL_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — Athena, духовний гід та ментор спільноти DAARION.city. Підтримуєш місію, цінності та зв\'язки між учасниками.",
)

# YAROMIR Configuration
YAROMIR_CONFIG = load_agent_config(
    agent_id="yaromir",
    name=os.getenv("YAROMIR_NAME", "Yaromir"),
    prompt_path=os.getenv(
        "YAROMIR_PROMPT_PATH",
        str(Path(__file__).parent / "yaromir_prompt.txt"),
    ),
    telegram_token_env="YAROMIR_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — Yaromir, стратег та наставник в екосистемі DAARION.city. Стратегія, наставництво, психологічна підтримка команди.",
)

# SOFIIA (Sophia) Configuration
SOFIIA_CONFIG = load_agent_config(
    agent_id="sofiia",
    name=os.getenv("SOFIIA_NAME", "Sophia"),
    prompt_path=os.getenv(
        "SOFIIA_PROMPT_PATH",
        str(Path(__file__).parent / "sofiia_prompt.txt"),
    ),
    telegram_token_env="SOFIIA_TELEGRAM_BOT_TOKEN",
    default_prompt="Ти — Sophia (Софія), Chief AI Architect та Technical Sovereign екосистеми DAARION.city. Координуєш R&D, архітектуру, безпеку та еволюцію платформи.",
)

# Registry of all agents (для легкого додавання нових агентів)
AGENT_REGISTRY: Dict[str, AgentConfig] = {
    "daarwizz": DAARWIZZ_CONFIG,
    "helion": HELION_CONFIG,
    "greenfood": GREENFOOD_CONFIG,
    "agromatrix": AGROMATRIX_CONFIG,
    "alateya": ALATEYA_CONFIG,
    "nutra": NUTRA_CONFIG,
    "druid": DRUID_CONFIG,
    "clan": CLAN_CONFIG,
    "eonarch": EONARCH_CONFIG,
    "senpai": SENPAI_CONFIG,
    "oneok": ONEOK_CONFIG,
    "soul": SOUL_CONFIG,
    "yaromir": YAROMIR_CONFIG,
    "sofiia": SOFIIA_CONFIG,
}
# 3. Створіть endpoint (опціонально, якщо потрібен окремий webhook):
#    @router.post("/new_agent/telegram/webhook")
#    async def new_agent_telegram_webhook(update: TelegramUpdate):
#        return await handle_telegram_webhook(NEW_AGENT_CONFIG, update)
#
# Новий агент автоматично отримає:
# - Обробку фото через Swapper vision-8b
# - Обробку PDF документів
# - Обробку голосових повідомлень (коли буде реалізовано)
# - RAG запити по документам
# - Memory context
    #    AGENT_REGISTRY["new_agent"] = NEW_AGENT_CONFIG
    # 3. Створіть endpoint (опціонально, якщо потрібен окремий webhook):


# Backward compatibility
DAARWIZZ_NAME = DAARWIZZ_CONFIG.name
DAARWIZZ_SYSTEM_PROMPT = DAARWIZZ_CONFIG.system_prompt
HELION_NAME = HELION_CONFIG.name
HELION_SYSTEM_PROMPT = HELION_CONFIG.system_prompt
GREENFOOD_NAME = GREENFOOD_CONFIG.name
GREENFOOD_SYSTEM_PROMPT = GREENFOOD_CONFIG.system_prompt


# ========================================
# Request Models
# ========================================

class TelegramUpdate(BaseModel):
    """Simplified Telegram update model"""
    update_id: Optional[int] = None
    message: Optional[Dict[str, Any]] = None
    channel_post: Optional[Dict[str, Any]] = None


# DRUID webhook endpoint
@router.post("/druid/telegram/webhook")
async def druid_telegram_webhook(update: TelegramUpdate):
    return await handle_telegram_webhook(DRUID_CONFIG, update)


# AGROMATRIX webhook endpoint

async def handle_stepan_message(update: TelegramUpdate, agent_config: AgentConfig) -> Dict[str, Any]:
    update_id = getattr(update, 'update_id', None) or update.update_id
    if update_id:
        if update_id in _PROCESSED_UPDATES:
            return {"ok": True, "status": "duplicate"}
        _PROCESSED_UPDATES[update_id] = _time.time()
        if len(_PROCESSED_UPDATES) > _DEDUP_MAX_SIZE:
            _dedup_cleanup()

    message = update.message or update.channel_post or {}
    text = message.get('text') or message.get('caption') or ''
    if not text:
        return {"ok": True, "status": "no_text"}

    user = message.get('from', {}) or {}
    chat = message.get('chat', {}) or {}
    user_id = str(user.get('id', ''))
    chat_id = str(chat.get('id', ''))

    # ops mode if operator
    ops_mode = False
    op_ids = [s.strip() for s in os.getenv('AGX_OPERATOR_IDS', '').split(',') if s.strip()]
    op_chat = os.getenv('AGX_OPERATOR_CHAT_ID', '').strip()
    if op_chat and chat_id == op_chat:
        ops_mode = True
    if user_id and user_id in op_ids:
        ops_mode = True

    trace_id = str(uuid.uuid4())

    # call Stepan directly
    try:
        sys.path.insert(0, str(Path('/opt/microdao-daarion')))
        from crews.agromatrix_crew.run import handle_message
        started = time.time()
        last_pending = _get_last_pending(chat_id)
        response_text = await asyncio.wait_for(
            asyncio.to_thread(handle_message, text, user_id, chat_id, trace_id, ops_mode, last_pending),
            timeout=25
        )
        duration_ms = int((time.time() - started) * 1000)
    except Exception as e:
        logger.error(f"Stepan handler error: {e}; trace_id={trace_id}")
        response_text = f"Помилка обробки. trace_id={trace_id}"
        duration_ms = 0

    # If JSON, try to show summary
    try:
        parsed = json.loads(response_text)
        summary = parsed.get('summary')
        if summary:
            response_text = summary
            if parsed.get('details'):
                response_text += "\n(details truncated)"
    except Exception:
        pass

    # chunk and send
    bot_token = agent_config.get_telegram_token()
    for chunk in _chunk_text(response_text, max_len=4096):
        await send_telegram_message(chat_id, chunk, bot_token=bot_token)

    logger.info(f"Stepan reply sent: trace_id={trace_id}, user_id={user_id}, chat_id={chat_id}, update_id={update_id}, duration_ms={duration_ms}")
    return {"ok": True}


@router.post("/agromatrix/telegram/webhook")
async def agromatrix_telegram_webhook(update: TelegramUpdate):
    # Check if this is an operator request (slash command or NL operator intent)
    message = (update.message or update.channel_post or {})
    msg_text = message.get('text') or message.get('caption') or ''
    user = message.get('from', {}) or {}
    chat = message.get('chat', {}) or {}
    user_id = str(user.get('id', ''))
    chat_id = str(chat.get('id', ''))

    is_slash = msg_text.strip().startswith('/')
    is_ops = False
    op_ids = [s.strip() for s in os.getenv('AGX_OPERATOR_IDS', '').split(',') if s.strip()]
    op_chat = os.getenv('AGX_OPERATOR_CHAT_ID', '').strip()
    if op_chat and chat_id == op_chat:
        is_ops = True
    if user_id and user_id in op_ids:
        is_ops = True

    # Operator NL or operator slash commands -> handle via Stepan handler.
    # Important: do NOT treat generic slash commands (/start, /agromatrix) as operator commands,
    # otherwise regular users will see "Недостатньо прав" or Stepan errors.
    operator_slash_cmds = {
        "whoami",
        "pending",
        "pending_show",
        "approve",
        "reject",
        "apply_dict",
        "pending_stats",
    }
    slash_cmd = ""
    if is_slash:
        try:
            slash_cmd = (msg_text.strip().split()[0].lstrip("/").strip().lower())
        except Exception:
            slash_cmd = ""
    is_operator_slash = bool(slash_cmd) and slash_cmd in operator_slash_cmds

    # Stepan handler currently depends on ChatOpenAI (OPENAI_API_KEY). If key is not configured,
    # never route production traffic there (avoid "Помилка обробки..." and webhook 5xx).
    stepan_enabled = bool(os.getenv("OPENAI_API_KEY", "").strip())
    if stepan_enabled and (is_ops or is_operator_slash):
        return await handle_stepan_message(update, AGROMATRIX_CONFIG)
    if (is_ops or is_operator_slash) and not stepan_enabled:
        logger.warning(
            "Stepan handler disabled (OPENAI_API_KEY missing); falling back to Router pipeline "
            f"for chat_id={chat_id}, user_id={user_id}, slash_cmd={slash_cmd!r}"
        )

    # General conversation -> standard Router pipeline (like all other agents)
    return await handle_telegram_webhook(AGROMATRIX_CONFIG, update)


# ALATEYA webhook endpoint
@router.post("/alateya/telegram/webhook")
async def alateya_telegram_webhook(update: TelegramUpdate):
    return await handle_telegram_webhook(ALATEYA_CONFIG, update)


# CLAN (Spirit) webhook endpoint
@router.post("/clan/telegram/webhook")
async def clan_telegram_webhook(update: TelegramUpdate):
    return await handle_telegram_webhook(CLAN_CONFIG, update)


# EONARCH webhook endpoint
@router.post("/eonarch/telegram/webhook")
async def eonarch_telegram_webhook(update: TelegramUpdate):
    return await handle_telegram_webhook(EONARCH_CONFIG, update)


# SENPAI (Gordon Senpai) webhook endpoint
@router.post("/senpai/telegram/webhook")
async def senpai_telegram_webhook(update: TelegramUpdate):
    return await handle_telegram_webhook(SENPAI_CONFIG, update)

# 1OK webhook endpoint
@router.post("/oneok/telegram/webhook")
async def oneok_telegram_webhook(update: TelegramUpdate):
    return await handle_telegram_webhook(ONEOK_CONFIG, update)


# SOUL / Athena webhook endpoint
@router.post("/soul/telegram/webhook")
async def soul_telegram_webhook(update: TelegramUpdate):
    return await handle_telegram_webhook(SOUL_CONFIG, update)


# YAROMIR webhook endpoint
@router.post("/yaromir/telegram/webhook")
async def yaromir_telegram_webhook(update: TelegramUpdate):
    return await handle_telegram_webhook(YAROMIR_CONFIG, update)


# SOFIIA (Sophia) webhook endpoint
@router.post("/sofiia/telegram/webhook")
async def sofiia_telegram_webhook(update: TelegramUpdate):
    return await handle_telegram_webhook(SOFIIA_CONFIG, update)


class DiscordMessage(BaseModel):
    """Simplified Discord message model"""
    content: Optional[str] = None
    author: Optional[Dict[str, Any]] = None
    channel_id: Optional[str] = None
    guild_id: Optional[str] = None


# ========================================
# DAO Mapping (temporary)
# ========================================

# Map agent_id to DAO ID
AGENT_TO_DAO = {
    "helion": "helion-dao",
    "greenfood": "greenfood-dao",
    "agromatrix": "agromatrix-dao",
    "nutra": "nutra-dao",
    "druid": "druid-dao",
    "daarwizz": "daarwizz-dao",
    "clan": "clan-dao",
    "alateya": "alateya-dao",
    "eonarch": "eonarch-dao",
    "senpai": "senpai-dao",
    "soul": "soul-dao",
    "yaromir": "yaromir-dao",
}

# Legacy: Map chat/channel ID to DAO ID
CHAT_TO_DAO = {
    "default": "daarion-dao",
}


def get_dao_id(chat_id: str, source: str, agent_id: str = None) -> str:
    """Get DAO ID from agent_id or chat ID"""
    if agent_id and agent_id in AGENT_TO_DAO:
        return AGENT_TO_DAO[agent_id]
    key = f"{source}:{chat_id}"
    return CHAT_TO_DAO.get(key, CHAT_TO_DAO["default"])


# ========================================
# Helper Functions
# ========================================

SERVICE_ACK_PREFIXES = (
    "📥 Імпортую",
    "📄 Обробляю",
    "Обробляю голосове",
    "🎤",
)


def is_service_response(text: str) -> bool:
    """Heuristic: визначає, чи відповідь є службовою (вітальна/ack)."""
    if not text:
        return True
    stripped = text.strip()
    if not stripped:
        return True
    if len(stripped) < 5:
        return True
    lower = stripped.lower()
    return any(lower.startswith(prefix.lower()) for prefix in SERVICE_ACK_PREFIXES)


def extract_bot_mentions(text: str) -> List[str]:
    """Витягує згадки інших ботів виду @NameBot."""
    if not text:
        return []
    mentions = []
    for token in text.split():
        if token.startswith("@") and token[1:].lower().endswith("bot"):
            mentions.append(token[1:])
    return mentions


def should_force_detailed_reply(text: str) -> bool:
    """Soft signal: user explicitly asks for details/long format."""
    if not text:
        return False
    lower = text.strip().lower()
    detail_markers = [
        "детально", "подробно", "розгорну", "розпиши", "по всіх пунктах",
        "step by step", "покроково", "з прикладами", "глибоко", "deep dive",
        "full", "повний розбір", "максимально детально",
    ]
    return any(m in lower for m in detail_markers)


def should_force_concise_reply(text: str) -> bool:
    """Soft concise mode by default, unless user asks for detailed answer."""
    if not text:
        return True

    stripped = text.strip()
    if not stripped:
        return True

    if should_force_detailed_reply(stripped):
        return False

    # Very long user request usually means they expect context-aware answer.
    if len(stripped) > 700:
        return False

    # For regular Q&A in chat keep first response concise by default.
    return True


COMPLEX_REASONING_KEYWORDS = [
    "стратег", "roadmap", "алгоритм", "architecture", "архітектур",
    "прогноз", "scenario", "модель", "аналіз", "побудуй", "plan", "дослідж",
    "симуляц", "forecast", "оптиміз", "розрахуй", "calculate", "predict"
]


def requires_complex_reasoning(text: str) -> bool:
    if not text:
        return False
    stripped = text.strip()
    if len(stripped) > 400:
        return True
    lower = stripped.lower()
    return any(keyword in lower for keyword in COMPLEX_REASONING_KEYWORDS)


LAST_RESPONSE_CACHE: Dict[Tuple[str, str], Dict[str, Any]] = {}
LAST_RESPONSE_TTL = float(os.getenv("TELEGRAM_LAST_RESPONSE_TTL", "15"))


def get_cached_response(agent_id: str, chat_id: str, text: str) -> Optional[str]:
    entry = LAST_RESPONSE_CACHE.get((agent_id, chat_id))
    if not entry:
        return None
    if entry["text"] == text and time.time() - entry["ts"] < LAST_RESPONSE_TTL:
        return entry["answer"]
    return None


def store_response_cache(agent_id: str, chat_id: str, text: str, answer: str) -> None:
    LAST_RESPONSE_CACHE[(agent_id, chat_id)] = {
        "text": text,
        "answer": answer,
        "ts": time.time(),
    }


def _resolve_stt_upload_url() -> str:
    """
    Повертає фінальний endpoint для STT.
    Swapper service використовує POST /stt з multipart file upload.
    """
    upload_override = os.getenv("STT_SERVICE_UPLOAD_URL")
    if upload_override:
        return upload_override.rstrip("/")
    
    base_url = os.getenv("STT_SERVICE_URL", "http://swapper-service:8890").rstrip("/")
    
    # Swapper endpoint is /stt (not /api/stt/upload)
    if base_url.endswith("/stt"):
        return base_url
    
    return f"{base_url}/stt"


# ========================================
# Helper Functions
# ========================================

async def get_telegram_file_path(file_id: str, bot_token: Optional[str] = None) -> Optional[str]:
    """
    Отримати шлях до файлу з Telegram API.
    
    Args:
        file_id: ID файлу з Telegram
        bot_token: Telegram bot token (якщо None, використовується TELEGRAM_BOT_TOKEN)
    
    Returns:
        Шлях до файлу або None
    """
    try:
        token = bot_token or os.getenv("TELEGRAM_BOT_TOKEN")
        if not token:
            logger.error("TELEGRAM_BOT_TOKEN not set")
            return None
        
        url = f"https://api.telegram.org/bot{token}/getFile"
        params = {"file_id": file_id}
        
        async with httpx.AsyncClient() as client:
            response = await client.get(url, params=params, timeout=10.0)
            response.raise_for_status()
            data = response.json()
            
            if data.get("ok"):
                return data.get("result", {}).get("file_path")
            return None
    except Exception as e:
        logger.error(f"Failed to get Telegram file path: {e}")
        return None


def format_qa_response(qa_list: list) -> str:
    """Форматувати список питань-відповідей для Telegram"""
    if not qa_list:
        return "Немає питань-відповідей."
    
    result = "📋 **Питання та відповіді:**\n\n"
    for i, qa in enumerate(qa_list, 1):
        question = qa.get("question", "") if isinstance(qa, dict) else getattr(qa, "question", "")
        answer = qa.get("answer", "") if isinstance(qa, dict) else getattr(qa, "answer", "")
        result += f"**{i}. {question}**\n{answer}\n\n"
    
    return result.strip()


def format_markdown_response(markdown: str) -> str:
    """Форматувати markdown відповідь для Telegram"""
    if len(markdown) > TELEGRAM_SAFE_LENGTH:
        return markdown[:TELEGRAM_SAFE_LENGTH] + "\n\n_... (відповідь обрізано)_"
    return markdown


def format_chunks_response(chunks: list) -> str:
    """Форматувати список чанків для Telegram"""
    if not chunks:
        return "Немає фрагментів."
    
    result = f"📄 **Знайдено {len(chunks)} фрагментів:**\n\n"
    for i, chunk in enumerate(chunks[:5], 1):  # Показуємо тільки перші 5
        text = chunk.get("text", "") if isinstance(chunk, dict) else str(chunk)
        if len(text) > 200:
            text = text[:200] + "..."
        result += f"**{i}.** {text}\n\n"
    
    if len(chunks) > 5:
        result += f"_... та ще {len(chunks) - 5} фрагментів_"
    
    return result.strip()


# ========================================
# Universal Message Processing Functions
# ========================================

async def process_photo(
    agent_config: AgentConfig,
    update: TelegramUpdate,
    chat_id: str,
    user_id: str,
    username: str,
    dao_id: str,
    photo: Dict[str, Any],
    caption_override: Optional[str] = None,
    bypass_media_gate: bool = False,
) -> Dict[str, Any]:
    """
    Універсальна функція для обробки фото для будь-якого агента.
    
    Args:
        agent_config: Конфігурація агента
        update: Telegram update об'єкт
        chat_id: ID чату
        user_id: ID користувача
        username: Ім'я користувача
        dao_id: ID DAO
        photo: Об'єкт фото з Telegram
    
    Returns:
        Dict з результатом обробки
    """
    # Telegram sends multiple sizes, get the largest one (last in array)
    photo_obj = photo[-1] if isinstance(photo, list) else photo
    file_id = photo_obj.get("file_id") if isinstance(photo_obj, dict) else None
    
    if not file_id:
        return {"ok": False, "error": "No file_id in photo"}
    
    logger.info(f"{agent_config.name}: Photo from {username} (tg:{user_id}), file_id: {file_id}")
    _set_recent_photo_context(agent_config.agent_id, chat_id, user_id, file_id)
    
    # Get caption for media question check
    caption = caption_override if caption_override is not None else ((update.message or {}).get("caption") or "")
    chat = (update.message or {}).get("chat", {})
    chat_type = chat.get("type", "private")
    is_private_chat = chat_type == "private"
    is_training = str(chat_id) in TRAINING_GROUP_IDS
    
    # BEHAVIOR POLICY v1: Media-no-comment
    # Check if photo has a question/request in caption
    if not bypass_media_gate and not is_private_chat and not is_training:
        has_question = detect_media_question(caption)
        if not has_question:
            logger.info(f"🔇 MEDIA-NO-COMMENT: Photo without question. Agent {agent_config.agent_id} NOT responding.")
            # Save to memory for context, but don't respond
            await memory_client.save_chat_turn(
                agent_id=agent_config.agent_id,
                team_id=dao_id,
                user_id=f"tg:{user_id}",
                message=f"[Photo: {file_id}] {caption}",
                response="",
                channel_id=chat_id,
                scope="short_term",
                save_agent_response=False,
                agent_metadata={
                    "media_no_comment": True,
                    "file_id": file_id,
                    "caption": caption,
                },
                username=username,
            )
            return {"ok": True, "skipped": True, "reason": "media_no_question"}
    
    try:
        # Get file path from Telegram
        telegram_token = agent_config.get_telegram_token()
        if not telegram_token:
            return {"ok": False, "error": f"Telegram token not configured for {agent_config.name}"}
        
        file_path = await get_telegram_file_path(file_id, telegram_token)
        if not file_path:
            raise HTTPException(status_code=400, detail="Failed to get file from Telegram")
        
        # Build file URL
        file_url = f"https://api.telegram.org/file/bot{telegram_token}/{file_path}"
        
        # Download and encode the image as base64 data URL for Router
        async with httpx.AsyncClient(timeout=60.0) as client:
            photo_resp = await client.get(file_url)
            photo_resp.raise_for_status()
            image_bytes = photo_resp.content
            content_type = photo_resp.headers.get("Content-Type", "")
        
        if not content_type or not content_type.startswith("image/"):
            content_type = "image/jpeg"
        
        encoded_image = base64.b64encode(image_bytes).decode("utf-8")
        data_url = f"data:{content_type};base64,{encoded_image}"
        
        logger.info(
            f"{agent_config.name}: Photo downloaded ({len(image_bytes)} bytes, content_type={content_type})"
        )
        
        # Send to Router with specialist_vision_8b model (Swapper)
        # IMPORTANT: Default prompt must request BRIEF description (2-3 sentences max)
        prompt = caption.strip() if caption else "Коротко (2-3 речення) скажи, що на цьому зображенні та яке його значення."
        router_request = {
            "message": f"{prompt}\n\n[Зображення передано окремо у context.images]",
            "mode": "chat",
            "agent": agent_config.agent_id,
            "payload": {
                "provider": "llm_specialist_vision_8b",
                "task_type": "vision_photo_analysis",
            },
            "metadata": {
                "source": "telegram",
                "dao_id": dao_id,
                "user_id": f"tg:{user_id}",
                "session_id": f"tg:{chat_id}:{dao_id}",
                "username": username,
                "chat_id": chat_id,
                "file_id": file_id,
                "file_url": file_url,
                "has_image": True,
                "provider": "llm_specialist_vision_8b",
                "use_llm": "specialist_vision_8b",
            },
            "context": {
                "agent_name": agent_config.name,
                "system_prompt": agent_config.system_prompt,
                "images": [data_url],
            },
        }
        
        # Send to Router
        logger.info(f"{agent_config.name}: Sending photo to Router with vision-8b (provider override)")
        response = await send_to_router(router_request)
        
        # Extract response
        if isinstance(response, dict) and response.get("ok"):
            answer_text = response.get("data", {}).get("text") or response.get("response", "")
            
            if answer_text:
                # Photo processed - send LLM response directly
                await send_telegram_message(
                    chat_id,
                    answer_text,  # No prefix, just the LLM response
                    telegram_token
                )
                
                # Save to memory for context
                await memory_client.save_chat_turn(
                    agent_id=agent_config.agent_id,
                    team_id=dao_id,
                    user_id=f"tg:{user_id}",
                    message=f"[Photo: {file_id}]",
                    response=answer_text,
                    channel_id=chat_id,
                    scope="short_term",
                    save_agent_response=not is_service_response(answer_text),
                    agent_metadata={"context": "photo"},
                    username=username,
                )
                
                return {"ok": True, "agent": agent_config.agent_id, "model": "specialist_vision_8b"}
            else:
                await send_telegram_message(
                    chat_id,
                    "Не вдалося коректно обробити фото. Спробуйте інше фото або додайте короткий опис, що саме перевірити.",
                    telegram_token
                )
                return {"ok": True, "handled": True, "reason": "vision_empty_response"}
        else:
            error_msg = response.get("error", "Unknown error") if isinstance(response, dict) else "Router error"
            logger.error(f"{agent_config.name}: Vision-8b error: {error_msg}")
            await send_telegram_message(
                chat_id,
                "Вибач, сталася помилка при обробці фото.",
                telegram_token
            )
            return {"ok": False, "error": error_msg}
        
    except Exception as e:
        logger.error(f"{agent_config.name}: Photo processing failed: {e}", exc_info=True)
        telegram_token = agent_config.get_telegram_token()
        await send_telegram_message(
            chat_id,
            "Вибач, сталася помилка при обробці фото.",
            telegram_token
        )
        return {"ok": False, "error": "Photo processing failed"}


async def process_document(
    agent_config: AgentConfig,
    update: TelegramUpdate,
    chat_id: str,
    user_id: str,
    username: str,
    dao_id: str,
    document: Dict[str, Any]
) -> Dict[str, Any]:
    """
    Універсальна функція для обробки документів для будь-якого агента.
    
    Args:
        agent_config: Конфігурація агента
        update: Telegram update об'єкт
        chat_id: ID чату
        user_id: ID користувача
        username: Ім'я користувача
        dao_id: ID DAO
        document: Об'єкт документа з Telegram
    
    Returns:
        Dict з результатом обробки
    """
    mime_type = document.get("mime_type", "")
    file_name = document.get("file_name", "")
    file_id = document.get("file_id")
    
    file_name_lower = file_name.lower()
    allowed_exts = {".pdf", ".docx", ".txt", ".md", ".csv", ".xlsx", ".zip"}
    is_allowed = any(file_name_lower.endswith(ext) for ext in allowed_exts)
    if mime_type == "application/pdf":
        is_allowed = True
    if mime_type in {
        "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
        "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
        "text/plain",
        "text/markdown",
        "text/csv",
        "application/zip",
    }:
        is_allowed = True
    
    if is_allowed and file_id:
        logger.info(f"{agent_config.name}: Document from {username} (tg:{user_id}), file_id: {file_id}, file_name: {file_name}")
        
        try:
            telegram_token = agent_config.get_telegram_token()
            if not telegram_token:
                return {"ok": False, "error": f"Telegram token not configured for {agent_config.name}"}
            
            file_path = await get_telegram_file_path(file_id, telegram_token)
            if not file_path:
                raise HTTPException(status_code=400, detail="Failed to get file from Telegram")
            
            file_url = f"https://api.telegram.org/file/bot{telegram_token}/{file_path}"
            
            session_id = f"telegram:{chat_id}"
            result = await parse_document(
                session_id=session_id,
                doc_url=file_url,
                file_name=file_name,
                dao_id=dao_id,
                user_id=f"tg:{user_id}",
                output_mode="qa_pairs",
                metadata={"username": username, "chat_id": chat_id}
            )
            
            if not result.success:
                await send_telegram_message(chat_id, f"Вибач, не вдалося обробити документ: {result.error}", telegram_token)
                return {"ok": False, "error": result.error}
            
            # Get document text for summary
            doc_text = result.markdown or ""
            if not doc_text and result.chunks_meta:
                chunks = result.chunks_meta.get("chunks", [])
                doc_text = "\n".join(chunks[:5]) if chunks else ""
            
            # Ask LLM to summarize the document (human-friendly)
            if doc_text:
                zip_hint = None
                if file_name_lower.endswith(".zip"):
                    zip_hint = _zip_read_summary(doc_text)
                summary_prompt = f"""Користувач надіслав документ "{file_name}". 
Ось його зміст (перші частини):

{doc_text[:3000]}

Дай коротке резюме цього документа в 2-3 реченнях:
- Про що цей документ?
- Яка його основна мета/тема?
- Що може бути корисним?

Відповідай українською, дружньо, без технічних термінів."""

                try:
                    summary_response = await send_to_router({
                        "message": summary_prompt,
                        "agent": agent_config.agent_id,
                        "context": {
                            "system_prompt": "Ти помічник який коротко пояснює зміст документів. Відповідай в 2-3 реченнях, дружньо і зрозуміло."
                        },
                        "metadata": {"source": "telegram", "task": "document_summary"}
                    })
                    
                    if isinstance(summary_response, dict) and summary_response.get("ok"):
                        answer_text = summary_response.get("response", "") or summary_response.get("data", {}).get("text", "")
                        if answer_text:
                            answer_text = f"📄 **{file_name}**\n\n{answer_text}"
                            if zip_hint:
                                answer_text = f"{zip_hint}\n\n{answer_text}"
                            answer_text += "\n\n_Що саме тебе цікавить у цьому документі?_"
                        else:
                            answer_text = f"📄 Отримав документ **{file_name}**. Що саме хочеш дізнатися з нього?"
                    else:
                        answer_text = f"📄 Отримав документ **{file_name}**. Про що саме хочеш запитати?"
                except Exception as e:
                    logger.warning(f"Failed to get document summary: {e}")
                    answer_text = f"📄 Отримав документ **{file_name}**. Що тебе цікавить?"
            else:
                answer_text = f"📄 Отримав документ **{file_name}**, але не вдалося прочитати текст. Можливо, це скановане зображення?"
            
            logger.info(f"{agent_config.name}: Document processed: {file_name}, doc_id={result.doc_id}")
            
            # === SAVE TO CHAT HISTORY (CRITICAL: so agent remembers the document) ===
            user_msg = f"[Документ: {file_name}] Надіслано документ"
            if update.message.get("caption"):
                user_msg = f"[Документ: {file_name}] {update.message.get('caption')}"
            await memory_client.save_chat_turn(
                agent_id=agent_config.agent_id,
                team_id=dao_id,
                user_id=f"tg:{user_id}",
                message=user_msg,
                response=answer_text,
                channel_id=chat_id,
                scope="short_term",
                save_agent_response=True,
                agent_metadata={"context": "document", "file_name": file_name, "doc_id": result.doc_id},
                username=username,
            )
            logger.info(f"{agent_config.name}: Document chat turn saved to memory: {file_name}")
            # === END SAVE TO CHAT HISTORY ===
            
            # === AUTO-INGEST: Store document in agent Qdrant _docs collection ===
            if doc_text:
                try:
                    import httpx as _httpx
                    router_url = os.getenv("ROUTER_URL", "http://router:8000")
                    async with _httpx.AsyncClient(timeout=60.0) as _client:
                        ingest_resp = await _client.post(
                            f"{router_url}/v1/documents/ingest",
                            json={
                                "agent_id": agent_config.agent_id,
                                "doc_id": result.doc_id,
                                "file_name": file_name,
                                "text": doc_text,
                                "dao_id": dao_id,
                                "user_id": f"tg:{user_id}"
                            }
                        )
                        ingest_data = ingest_resp.json()
                        if ingest_data.get("ok"):
                            logger.info(f"{agent_config.name}: Document ingested to Qdrant: {ingest_data.get('chunks_stored', 0)} chunks")
                        else:
                            logger.warning(f"{agent_config.name}: Document ingest failed: {ingest_data.get('error')}")
                except Exception as ingest_err:
                    logger.warning(f"{agent_config.name}: Document auto-ingest error: {ingest_err}")
            # === END AUTO-INGEST ===
            
            await send_telegram_message(chat_id, answer_text, telegram_token)
            return {"ok": True, "agent": "parser", "mode": "doc_parse", "doc_id": result.doc_id}
            
        except Exception as e:
            logger.error(f"{agent_config.name}: Document processing failed: {e}", exc_info=True)
            telegram_token = agent_config.get_telegram_token()
            await send_telegram_message(chat_id, "Вибач, не вдалося обробити документ. Переконайся, що файл не пошкоджений.", telegram_token)
            return {"ok": False, "error": "Document processing failed"}
    elif document and not is_allowed:
        telegram_token = agent_config.get_telegram_token()
        await send_telegram_message(
            chat_id,
            "Наразі підтримуються формати: PDF, DOCX, TXT, MD, CSV, XLSX, ZIP.",
            telegram_token,
        )
        return {"ok": False, "error": "Unsupported document type"}
    
    return {"ok": False, "error": "No document to process"}


async def process_voice(
    agent_config: AgentConfig,
    update: TelegramUpdate,
    chat_id: str,
    user_id: str,
    username: str,
    dao_id: str,
    media_obj: Dict[str, Any]
) -> Dict[str, Any]:
    """
    Універсальна функція для обробки голосових повідомлень для будь-якого агента.
    Використовує STT Service для розпізнавання мовлення.
    
    Args:
        agent_config: Конфігурація агента
        update: Telegram update об'єкт
        chat_id: ID чату
        user_id: ID користувача
        username: Ім'я користувача
        dao_id: ID DAO
        media_obj: Об'єкт голосового повідомлення з Telegram
    
    Returns:
        Dict з результатом обробки та розпізнаним текстом
    """
    file_id = media_obj.get("file_id") if media_obj else None
    
    if not file_id:
        return {"ok": False, "error": "No file_id in voice/audio/video_note"}
    
    logger.info(f"{agent_config.name}: Voice message from {username} (tg:{user_id}), file_id: {file_id}")
    
    try:
        telegram_token = agent_config.get_telegram_token()
        if not telegram_token:
            return {"ok": False, "error": f"Telegram token not configured for {agent_config.name}"}
        
        # Отримуємо файл з Telegram
        file_path = await get_telegram_file_path(file_id, telegram_token)
        if not file_path:
            raise HTTPException(status_code=400, detail="Failed to get file from Telegram")
        
        # Завантажуємо файл
        file_url = f"https://api.telegram.org/file/bot{telegram_token}/{file_path}"
        async with httpx.AsyncClient(timeout=30.0) as client:
            file_resp = await client.get(file_url)
            file_resp.raise_for_status()
            audio_bytes = file_resp.content
        
        # Відправляємо в STT-сервіс
        stt_upload_url = _resolve_stt_upload_url()
        mime_type = media_obj.get("mime_type") if isinstance(media_obj, dict) else None
        files = {
            "file": (
                "voice.ogg",
                audio_bytes,
                mime_type or "audio/ogg",
            )
        }
        # Swapper /stt expects: file (multipart), model (form), language (form)
        form_data = {
            "model": "whisper-small",
            "task": "transcribe",
        }
        
        logger.info(f"{agent_config.name}: Sending voice to STT endpoint {stt_upload_url}")
        async with httpx.AsyncClient(timeout=90.0) as client:
            stt_resp = await client.post(stt_upload_url, files=files, data=form_data)
            stt_resp.raise_for_status()
            stt_data = stt_resp.json()
            text = stt_data.get("text", "")
        
        if not text:
            await send_telegram_message(
                chat_id,
                "Вибач, не вдалося розпізнати голосове повідомлення. Спробуй надіслати текстом.",
                telegram_token
            )
            return {"ok": False, "error": "STT returned empty text"}
        
        logger.info(f"{agent_config.name}: STT result: {text[:100]}...")
        
        # Повертаємо розпізнаний текст для подальшої обробки
        return {"ok": True, "text": text, "agent": agent_config.agent_id, "mode": "voice_stt"}
        
    except Exception as e:
        logger.error(f"{agent_config.name}: Voice processing failed: {e}", exc_info=True)
        telegram_token = agent_config.get_telegram_token()
        await send_telegram_message(
            chat_id,
            "Вибач, не вдалося розпізнати голосове повідомлення. Спробуй надіслати текстом.",
            telegram_token
        )
        return {"ok": False, "error": "Voice processing failed"}


# ========================================
# Universal Telegram Webhook Handler
# ========================================

# === UPDATE DEDUPLICATION ===
import time as _time
_PROCESSED_UPDATES: Dict[int, float] = {}  # update_id -> timestamp
_DEDUP_MAX_SIZE = 2000
_DEDUP_TTL = 300  # 5 minutes

def _dedup_cleanup():
    """Remove old entries from dedup cache."""
    now = _time.time()
    expired = [uid for uid, ts in _PROCESSED_UPDATES.items() if now - ts > _DEDUP_TTL]
    for uid in expired:
        del _PROCESSED_UPDATES[uid]
# === END DEDUPLICATION ===


async def handle_telegram_webhook(
    agent_config: AgentConfig,
    update: TelegramUpdate
) -> Dict[str, Any]:
    """
    Універсальна функція для обробки Telegram webhook для будь-якого агента.
    
    Args:
        agent_config: Конфігурація агента
        update: Telegram update об'єкт
    
    Returns:
        Dict з результатом обробки
    """
    # Allow updates without message if they contain photo/voice
    # The actual message validation happens after multimodal checks
    # === DEDUP CHECK ===
    if update.update_id:
        if update.update_id in _PROCESSED_UPDATES:
            logger.info(f"🔄 Skipping duplicate update_id={update.update_id} for {agent_config.name}")
            return {"status": "ok", "skipped": "duplicate_update"}
        _PROCESSED_UPDATES[update.update_id] = _time.time()
        if len(_PROCESSED_UPDATES) > _DEDUP_MAX_SIZE:
            _dedup_cleanup()
    # === END DEDUP CHECK ===

    if not update.message:
        if update.channel_post:
            update.message = update.channel_post
        else:
            return {"status": "ok", "skipped": "no_message"}
    
    # Extract message details
    from_user = update.message.get("from", {})
    if not from_user:
        from_user = update.message.get("sender_chat", {})
    chat = update.message.get("chat", {})
    
    user_id = str(from_user.get("id", "unknown"))
    chat_id = str(chat.get("id", "unknown"))
    username = from_user.get("username", "")
    first_name = from_user.get("first_name")
    last_name = from_user.get("last_name")
    is_sender_bot = bool(from_user.get("is_bot") or (username and username.lower().endswith("bot")))
    
    # Get DAO ID for this chat
    dao_id = get_dao_id(chat_id, "telegram", agent_id=agent_config.agent_id)
    
    initial_preferred_lang = resolve_preferred_language(
        chat_id=chat_id,
        user_id=user_id,
        text=update.message.get("text", ""),
        telegram_lang_code=from_user.get("language_code"),
    )

    # Оновлюємо факти про користувача/агента для побудови графу пам'яті
    asyncio.create_task(
        memory_client.upsert_fact(
            user_id=f"tg:{user_id}",
            fact_key="profile",
            fact_value_json={
                "username": username,
                "first_name": first_name,
                "last_name": last_name,
                "language_code": from_user.get("language_code"),
                "preferred_language": initial_preferred_lang,
                "is_bot": is_sender_bot,
            },
            team_id=dao_id,
        )
    )
    
    telegram_token = agent_config.get_telegram_token()
    if not telegram_token:
        raise HTTPException(status_code=500, detail=f"Telegram token not configured for {agent_config.name}")
    
    # === REPLY-TO-AGENT DETECTION ===
    # If user replies to a bot message → treat as direct mention (SOWA Priority 3)
    is_reply_to_agent = False
    reply_to_message = update.message.get("reply_to_message")
    if reply_to_message:
        reply_from = reply_to_message.get("from", {})
        if reply_from.get("is_bot"):
            # Verify it's THIS agent's bot (bot_id = first part of token)
            bot_id = telegram_token.split(":")[0] if telegram_token else None
            reply_from_id = str(reply_from.get("id", ""))
            if bot_id and reply_from_id == bot_id:
                is_reply_to_agent = True
                logger.info(
                    f"↩️ {agent_config.name}: Reply-to-agent detected "
                    f"(user {username} replied to bot msg {reply_to_message.get('message_id', '?')})"
                )

    # === BRAND COMMANDS (experimental, disabled by default) ===
    # ~290 lines of brand/presentation command handlers.
    # This code was NEVER active in production (was trapped inside wrong indent block).
    # Quarantined behind feature flag. Enable with: ENABLE_BRAND_COMMANDS=true
    # See: /бренд, /бренд_інтейк, /бренд_тема, /презентація, /job_статус
    _brand_commands_enabled = os.environ.get("ENABLE_BRAND_COMMANDS", "").lower() in ("1", "true", "yes")
    if _brand_commands_enabled:
        text = update.message.get("text", "")

        # Simple brand commands (Ukrainian)
        if text and text.strip().startswith("/бренд"):
            parts = text.strip().split(maxsplit=2)
            command = parts[0].lower()
            if command == "/бренд":
                await send_telegram_message(
                    chat_id,
                    "🧩 **Команди бренду**\n\n"
                    "• `/бренд_інтейк <url|текст>` — зберегти джерело\n"
                    "• `/бренд_тема <brand_id> [версія]` — опублікувати базову тему\n"
                    "• `/бренд_останнє <brand_id>` — показати останню тему\n"
                    "• `/бренд_показати <brand_id> <версія>` — показати конкретну тему\n"
                    "• `/презентація <brand_id> <версія> <JSON SlideSpec>` — рендер презентації\n"
                    "• `/презентація_статус <job_id>` — статус рендера\n"
                    "• `/презентація_файл <artifact_id> [pptx|pdf]` — файл\n"
                    "• `/job_статус <job_id>` — універсальний статус",
                    telegram_token
                )
                return {"ok": True, "action": "brand_help"}

            if command == "/бренд_інтейк":
                if len(parts) < 2:
                    await send_telegram_message(
                        chat_id,
                        "❗ Вкажи URL або текст: `/бренд_інтейк <url|текст>`",
                        telegram_token
                    )
                    return {"ok": True, "action": "brand_intake_help"}
                source_value = parts[1] if len(parts) == 2 else f"{parts[1]} {parts[2]}"
                source_type = "url" if source_value.startswith("http") else "text"
                intake_payload = {
                    "source_type": source_type,
                    "text": source_value if source_type == "text" else None,
                    "url": source_value if source_type == "url" else None,
                    "agent_id": agent_config.agent_id,
                    "workspace_id": dao_id,
                    "project_id": dao_id,
                    "tags": ["telegram"]
                }
                result = await _brand_intake_request(intake_payload)
                attribution = result.get("attribution", {})
                await send_telegram_message(
                    chat_id,
                    "✅ **Джерело збережено**\n\n"
                    f"ID: `{result.get('id')}`\n"
                    f"Статус: `{attribution.get('status')}`\n"
                    f"Бренд: `{attribution.get('brand_id')}`\n"
                    f"Впевненість: `{attribution.get('confidence')}`",
                    telegram_token
                )
                return {"ok": True, "action": "brand_intake"}

            if command == "/бренд_тема":
                if len(parts) < 2:
                    await send_telegram_message(
                        chat_id,
                        "❗ Вкажи brand_id: `/бренд_тема <brand_id> [версія]`",
                        telegram_token
                    )
                    return {"ok": True, "action": "brand_theme_help"}
                brand_id = parts[1]
                theme_version = None
                if len(parts) == 3:
                    theme_version = parts[2]
                theme = _default_theme_payload(brand_id)
                published = await _brand_publish_theme(brand_id, theme, theme_version)
                await send_telegram_message(
                    chat_id,
                    "✅ **Тему опубліковано**\n\n"
                    f"Бренд: `{published.get('brand_id')}`\n"
                    f"Версія: `{published.get('theme_version')}`",
                    telegram_token
                )
                return {"ok": True, "action": "brand_publish"}

            if command == "/бренд_останнє":
                if len(parts) < 2:
                    await send_telegram_message(
                        chat_id,
                        "❗ Вкажи brand_id: `/бренд_останнє <brand_id>`",
                        telegram_token
                    )
                    return {"ok": True, "action": "brand_latest_help"}
                brand_id = parts[1]
                data = await _brand_get_latest(brand_id)
                await send_telegram_message(
                    chat_id,
                    "📌 **Остання тема**\n\n"
                    f"Бренд: `{data.get('brand_id')}`\n"
                    f"Версія: `{data.get('theme_version')}`",
                    telegram_token
                )
                return {"ok": True, "action": "brand_latest"}

            if command == "/бренд_показати":
                if len(parts) < 3:
                    await send_telegram_message(
                        chat_id,
                        "❗ Формат: `/бренд_показати <brand_id> <версія>`",
                        telegram_token
                    )
                    return {"ok": True, "action": "brand_show_help"}
                brand_id = parts[1]
                theme_version = parts[2]
                data = await _brand_get_theme(brand_id, theme_version)
                await send_telegram_message(
                    chat_id,
                    "📎 **Тема**\n\n"
                    f"Бренд: `{data.get('brand_id')}`\n"
                    f"Версія: `{data.get('theme_version')}`",
                    telegram_token
                )
                return {"ok": True, "action": "brand_show"}

        # Brand hint on keyword mention (non-command)
        if text and "бренд" in text.lower():
            await send_telegram_message(
                chat_id,
                "🧩 **Команди бренду**\n\n"
                "• `/бренд_інтейк <url|текст>` — зберегти джерело\n"
                "• `/бренд_тема <brand_id> [версія]` — опублікувати базову тему\n"
                "• `/бренд_останнє <brand_id>` — показати останню тему\n"
                "• `/бренд_показати <brand_id> <версія>` — показати конкретну тему\n"
                "• `/презентація <brand_id> <версія> <JSON SlideSpec>` — рендер презентації\n"
                "• `/презентація_статус <job_id>` — статус рендера\n"
                "• `/презентація_файл <artifact_id> [pptx|pdf]` — файл\n"
                "• `/job_статус <job_id>` — універсальний статус",
                telegram_token
            )
            return {"ok": True, "action": "brand_hint"}

        # Job status command (universal)
        if text and text.strip().startswith("/job_статус"):
            parts = text.strip().split(maxsplit=1)
            if len(parts) < 2:
                await send_telegram_message(
                    chat_id,
                    "❗ Формат: `/job_статус <job_id>`",
                    telegram_token
                )
                return {"ok": True, "action": "job_status_help"}
            job_id = parts[1].strip()
            job = await _artifact_job_status(job_id)
            message = _format_job_status_message(job, job_id)
            await send_telegram_message(chat_id, message, telegram_token)
            return {"ok": True, "action": "job_status"}

        # Presentation status command (alias)
        if text and text.strip().startswith("/презентація_статус"):
            parts = text.strip().split(maxsplit=1)
            if len(parts) < 2:
                await send_telegram_message(
                    chat_id,
                    "❗ Формат: `/презентація_статус <job_id>` або `/job_статус <job_id>`",
                    telegram_token
                )
                return {"ok": True, "action": "presentation_status_help"}
            job_id = parts[1].strip()
            job = await _artifact_job_status(job_id)
            message = _format_job_status_message(job, job_id)
            await send_telegram_message(chat_id, message, telegram_token)
            return {"ok": True, "action": "presentation_status"}

        # Presentation file command
        if text and text.strip().startswith("/презентація_файл"):
            parts = text.strip().split(maxsplit=2)
            if len(parts) < 2:
                await send_telegram_message(
                    chat_id,
                    "❗ Формат: `/презентація_файл <artifact_id> [pptx|pdf]`",
                    telegram_token
                )
                return {"ok": True, "action": "presentation_file_help"}
            artifact_id = parts[1].strip()
            fmt = parts[2].strip().lower() if len(parts) > 2 else "pptx"

            artifact = await _artifact_get(artifact_id)
            acl_ref = artifact.get("acl_ref")
            if not _can_access_artifact(acl_ref, dao_id, f"tg:{user_id}"):
                await send_telegram_message(chat_id, "⛔ Немає доступу до цього файлу.", telegram_token)
                return {"ok": True, "action": "presentation_file_denied"}

            try:
                download = await _artifact_download(artifact_id, fmt)
                logger.info(
                    "artifact.downloaded artifact_id=%s user=%s format=%s",
                    artifact_id,
                    user_id,
                    fmt,
                )
                await send_telegram_message(
                    chat_id,
                    f"📎 **Файл готовий** ({fmt})\n{download.get('url')}",
                    telegram_token
                )
                return {"ok": True, "action": "presentation_file"}
            except HTTPException as e:
                if fmt == "pdf" and e.status_code == 404:
                    versions = await _artifact_versions(artifact_id)
                    pptx_version_id = None
                    for item in versions.get("items", []):
                        if item.get("mime") == "application/vnd.openxmlformats-officedocument.presentationml.presentation":
                            pptx_version_id = item.get("id")
                            break
                    if not pptx_version_id:
                        await send_telegram_message(
                            chat_id,
                            "❗ PPTX ще не готовий, PDF теж недоступний.",
                            telegram_token
                        )
                        return {"ok": True, "action": "presentation_pdf_missing"}
                    job = await _artifact_create_job(artifact_id, "render_pdf", pptx_version_id)
                    await send_telegram_message(
                        chat_id,
                        "⏳ PDF в черзі на рендер.\n"
                        f"Job ID: `{job.get('job_id')}`\n"
                        "Спробуй `/презентація_статус <job_id>` трохи пізніше.",
                        telegram_token
                    )
                    return {"ok": True, "action": "presentation_pdf_queued"}
                raise

        # Presentation render command (JSON SlideSpec)
        if text and text.strip().startswith("/презентація"):
            parts = text.strip().split(maxsplit=3)
            if len(parts) < 4:
                await send_telegram_message(
                    chat_id,
                    "❗ Формат:\n"
                    "`/презентація <brand_id> <версія> <JSON SlideSpec>`\n"
                    "або простий формат:\n"
                    "`/презентація <brand_id> <версія> Назва;Слайд 1;Слайд 2;Слайд 3`\n\n"
                    "Приклад:\n"
                    "`/презентація energyunion v1.0.0 {\"meta\":{\"title\":\"Pitch\",\"brand_id\":\"energyunion\",\"theme_version\":\"v1.0.0\",\"language\":\"uk\"},\"slides\":[{\"type\":\"title\",\"title\":\"Energy Union\"}]}`",
                    telegram_token
                )
                return {"ok": True, "action": "presentation_help"}
            brand_id = parts[1]
            theme_version = parts[2]
            slidespec_raw = parts[3]
            slidespec = None
            if slidespec_raw.strip().startswith("{"):
                try:
                    slidespec = json.loads(slidespec_raw)
                except json.JSONDecodeError:
                    await send_telegram_message(
                        chat_id,
                        "❗ Не вдалося прочитати JSON SlideSpec. Перевір формат.",
                        telegram_token
                    )
                    return {"ok": True, "action": "presentation_bad_json"}
            else:
                parts_simple = [p.strip() for p in slidespec_raw.split(";") if p.strip()]
                if not parts_simple:
                    await send_telegram_message(
                        chat_id,
                        "❗ Порожній список слайдів. Додай хоча б назву.",
                        telegram_token
                    )
                    return {"ok": True, "action": "presentation_empty"}
                title = parts_simple[0]
                slides = [{"type": "title", "title": title}]
                for item in parts_simple[1:]:
                    slides.append({
                        "type": "bullets",
                        "title": item,
                        "blocks": [{"kind": "bullets", "items": [item]}],
                    })
                slidespec = {
                    "meta": {
                        "title": title,
                        "brand_id": brand_id,
                        "theme_version": theme_version,
                        "language": "uk",
                    },
                    "slides": slides,
                }

            render_result = await _presentation_render(slidespec, brand_id, theme_version)
            await send_telegram_message(
                chat_id,
                "✅ **Запит на рендер прийнято**\n\n"
                f"Artifact ID: `{render_result.get('artifact_id')}`\n"
                f"Job ID: `{render_result.get('job_id')}`\n"
                f"Status URL: `{render_result.get('status_url')}`",
                telegram_token
            )
            return {"ok": True, "action": "presentation_render"}

    # Check for /ingest command
    text = update.message.get("text", "")
    if text and text.strip().startswith("/ingest"):
        session_id = f"telegram:{chat_id}"
        
        # Check if there's a document in the message
        document = update.message.get("document")
        if document:
            mime_type = document.get("mime_type", "")
            file_name = document.get("file_name", "")
            file_id = document.get("file_id")
            
            if file_id:
                try:
                    file_path = await get_telegram_file_path(file_id, telegram_token)
                    if file_path:
                        file_url = f"https://api.telegram.org/file/bot{telegram_token}/{file_path}"
                        artifact = None
                        job = None
                        try:
                            artifact = await _artifact_create({
                                "type": "doc",
                                "title": file_name,
                                "brand_id": dao_id,
                                "project_id": dao_id,
                                "acl_ref": f"brand:{dao_id}:public" if dao_id else "public",
                                "created_by": f"tg:{user_id}",
                            })
                            version = await _artifact_add_version_from_url(
                                artifact["artifact_id"],
                                {
                                    "url": file_url,
                                    "mime": mime_type or "application/octet-stream",
                                    "label": "source",
                                    "meta_json": {
                                        "file_name": file_name,
                                        "dao_id": dao_id,
                                        "user_id": f"tg:{user_id}",
                                    },
                                },
                            )
                            job = await _artifact_create_job(
                                artifact["artifact_id"],
                                "index_doc",
                                version["version_id"],
                            )
                        except Exception as e:
                            logger.warning(f"Artifact doc registry failed: {e}")
                        await send_telegram_message(
                            chat_id,
                            f"✅ **Документ прийнято**\n\n"
                            f"📁 DAO: {dao_id}\n"
                            f"🧾 Artifact ID: `{artifact.get('artifact_id') if artifact else 'n/a'}`\n"
                            f"🧩 Job ID: `{job.get('job_id') if job else 'n/a'}`\n\n"
                            f"Індексація виконується асинхронно. "
                            f"Перевір статус: `/job_статус <job_id>`",
                            telegram_token
                        )
                        return {"ok": True, "artifact_id": artifact.get("artifact_id") if artifact else None, "job_id": job.get("job_id") if job else None}
                except Exception as e:
                    logger.error(f"{agent_config.name}: Ingest failed: {e}", exc_info=True)
                    await send_telegram_message(chat_id, "Вибач, не вдалося імпортувати документ.", telegram_token)
                    return {"ok": False, "error": "Ingest failed"}
        
        await send_telegram_message(chat_id, "Спочатку надішли документ, а потім використай /ingest", telegram_token)
        return {"ok": False, "error": "No document for ingest"}
    
    # Check for /link command - Account Linking (Telegram ↔ Energy Union)
    if text and text.strip().startswith("/link"):
        parts = text.strip().split(maxsplit=1)
        if len(parts) < 2:
            await send_telegram_message(
                chat_id,
                "🔗 **Зв'язування акаунта**\n\n"
                "Щоб зв'язати Telegram з акаунтом Energy Union:\n"
                "1. Отримай код у кабінеті Energy Union\n"
                "2. Надішли: `/link <код>`\n\n"
                "Приклад: `/link ABC123XYZ`",
                telegram_token
            )
            return {"ok": True, "action": "link_help"}
        
        link_code = parts[1].strip()
        
        # Call PostgreSQL function to complete linking
        try:
            import asyncpg
            pg_conn = await asyncpg.connect(
                host=os.getenv("POSTGRES_HOST", "dagi-postgres"),
                port=int(os.getenv("POSTGRES_PORT", "5432")),
                user=os.getenv("POSTGRES_USER", "daarion"),
                password=os.getenv("POSTGRES_PASSWORD", "DaarionDB2026!"),
                database=os.getenv("POSTGRES_DB", "daarion_main")
            )
            
            result = await pg_conn.fetchrow(
                "SELECT * FROM complete_account_link($1, $2, $3, $4, $5)",
                link_code,
                int(user_id),
                username,
                first_name,
                last_name
            )
            await pg_conn.close()
            
            if result and result['success']:
                await send_telegram_message(
                    chat_id,
                    "✅ **Акаунт успішно зв'язано!**\n\n"
                    "Тепер Helion бачить твою історію взаємодій "
                    "з платформою Energy Union.\n\n"
                    "Твої розмови в різних чатах тепер пов'язані "
                    "з твоїм єдиним акаунтом.",
                    telegram_token
                )
                logger.info(f"Account linked: telegram_user_id={user_id}, account_id={result['account_id']}")
                return {"ok": True, "action": "account_linked", "account_id": str(result['account_id'])}
            else:
                error_msg = result['error_message'] if result else "Невідома помилка"
                error_text = {
                    "Invalid or expired code": "Код недійсний або прострочений",
                    "Telegram account already linked": "Telegram вже зв'язано з іншим акаунтом",
                    "Code not found": "Код не знайдено"
                }.get(error_msg, error_msg)
                
                await send_telegram_message(
                    chat_id,
                    f"❌ **Не вдалося зв'язати акаунт**\n\n"
                    f"Причина: {error_text}\n\n"
                    "Спробуй отримати новий код у кабінеті Energy Union.",
                    telegram_token
                )
                return {"ok": False, "error": error_msg}
                
        except Exception as e:
            logger.error(f"Account linking failed: {e}", exc_info=True)
            await send_telegram_message(
                chat_id,
                "❌ Помилка зв'язування акаунта. Спробуй пізніше.",
                telegram_token
            )
            return {"ok": False, "error": str(e)}
    
    # Check for /unlink command
    if text and text.strip().startswith("/unlink"):
        try:
            import asyncpg
            pg_conn = await asyncpg.connect(
                host=os.getenv("POSTGRES_HOST", "dagi-postgres"),
                port=int(os.getenv("POSTGRES_PORT", "5432")),
                user=os.getenv("POSTGRES_USER", "daarion"),
                password=os.getenv("POSTGRES_PASSWORD", "DaarionDB2026!"),
                database=os.getenv("POSTGRES_DB", "daarion_main")
            )
            
            result = await pg_conn.execute(
                """
                UPDATE account_links 
                SET status = 'revoked', 
                    revoked_at = NOW(), 
                    revoked_reason = 'User requested via /unlink'
                WHERE telegram_user_id = $1 AND status = 'active'
                """,
                int(user_id)
            )
            await pg_conn.close()
            
            await send_telegram_message(
                chat_id,
                "✅ **Зв'язок з акаунтом видалено**\n\n"
                "Helion більше не бачить твою історію.\n"
                "Ти можеш повторно зв'язати акаунт командою `/link`.",
                telegram_token
            )
            return {"ok": True, "action": "account_unlinked"}
            
        except Exception as e:
            logger.error(f"Account unlinking failed: {e}", exc_info=True)
            await send_telegram_message(
                chat_id,
                "❌ Помилка видалення зв'язку. Спробуй пізніше.",
                telegram_token
            )
            return {"ok": False, "error": str(e)}
    
    # Check for /status command - Show linking status
    if text and text.strip().startswith("/status"):
        try:
            import asyncpg
            pg_conn = await asyncpg.connect(
                host=os.getenv("POSTGRES_HOST", "dagi-postgres"),
                port=int(os.getenv("POSTGRES_PORT", "5432")),
                user=os.getenv("POSTGRES_USER", "daarion"),
                password=os.getenv("POSTGRES_PASSWORD", "DaarionDB2026!"),
                database=os.getenv("POSTGRES_DB", "daarion_main")
            )
            
            link = await pg_conn.fetchrow(
                """
                SELECT account_id, linked_at, status 
                FROM account_links 
                WHERE telegram_user_id = $1 AND status = 'active'
                """,
                int(user_id)
            )
            await pg_conn.close()
            
            if link:
                linked_date = link['linked_at'].strftime("%d.%m.%Y %H:%M")
                await send_telegram_message(
                    chat_id,
                    f"✅ **Акаунт зв'язано**\n\n"
                    f"📅 Дата: {linked_date}\n"
                    f"🔗 Статус: активний\n\n"
                    f"Helion бачить твою історію взаємодій.",
                    telegram_token
                )
            else:
                await send_telegram_message(
                    chat_id,
                    "❌ **Акаунт не зв'язано**\n\n"
                    "Використай `/link <код>` щоб зв'язати.\n"
                    "Код можна отримати в кабінеті Energy Union.",
                    telegram_token
                )
            return {"ok": True, "action": "status_checked", "linked": bool(link)}
            
        except Exception as e:
            logger.error(f"Status check failed: {e}", exc_info=True)
            return {"ok": False, "error": str(e)}
    
    # Check if it's a document
    document = update.message.get("document")
    if document:
        result = await process_document(
            agent_config, update, chat_id, user_id, username, dao_id, document
        )
        if result.get("ok"):
            return result
    
    # Check if it's a photo
    photo = update.message.get("photo")
    if photo:
        result = await process_photo(
            agent_config, update, chat_id, user_id, username, dao_id, photo
        )
        return result
    
    # Check if it's a voice message
    voice = update.message.get("voice")
    audio = update.message.get("audio")
    video_note = update.message.get("video_note")
    
    text = ""
    if voice or audio or video_note:
        media_obj = voice or audio or video_note
        result = await process_voice(
            agent_config, update, chat_id, user_id, username, dao_id, media_obj
        )
        if result.get("ok") and result.get("text"):
            # Отримали розпізнаний текст, продовжуємо обробку як текстове повідомлення
            text = result.get("text")
        elif result.get("ok"):
            # STT успішний, але текст порожній
            return result
        else:
            # Помилка STT
            return result
    
    # Get message text (якщо не було голосового повідомлення)
    if not text:
        text = update.message.get("text", "")
        caption = update.message.get("caption", "")

    # If user asks about a recently sent photo, run vision on cached photo file_id.
    if text and _looks_like_photo_followup(text):
        recent_file_id = _get_recent_photo_file_id(agent_config.agent_id, chat_id, user_id)
        if recent_file_id:
            logger.info(
                f"{agent_config.name}: Detected follow-up photo question; using cached file_id={recent_file_id}"
            )
            followup_result = await process_photo(
                agent_config=agent_config,
                update=update,
                chat_id=chat_id,
                user_id=user_id,
                username=username,
                dao_id=dao_id,
                photo={"file_id": recent_file_id},
                caption_override=text,
                bypass_media_gate=True,
            )
            return followup_result
        
        if not text and not caption:
            # Check for unsupported message types and silently ignore
            unsupported_types = ["sticker", "animation", "video_note", "contact", "location", 
                                "venue", "poll", "dice", "game", "new_chat_members", 
                                "left_chat_member", "new_chat_title", "new_chat_photo",
                                "delete_chat_photo", "pinned_message", "message_auto_delete_timer_changed"]
            for msg_type in unsupported_types:
                if update.message.get(msg_type):
                    logger.debug(f"Ignoring unsupported message type: {msg_type}")
                    return {"ok": True, "ignored": True, "reason": f"Unsupported message type: {msg_type}"}
            
            # If no supported content found, return silently
            logger.debug(f"Message without processable content from user {user_id}")
            return {"ok": True, "ignored": True, "reason": "No processable content"}
        
        # Use caption if text is empty (for photos with captions that weren't processed)
        if not text and caption:
            text = caption
    
    logger.info(f"{agent_config.name} Telegram message from {username} (tg:{user_id}) in chat {chat_id}: {text[:50]}")
    mentioned_bots = extract_bot_mentions(text)
    needs_complex_reasoning = requires_complex_reasoning(text)
    
    cached_answer = get_cached_response(agent_config.agent_id, chat_id, text)
    if cached_answer:
        await send_telegram_message(chat_id, cached_answer, telegram_token)
        await memory_client.save_chat_turn(
            agent_id=agent_config.agent_id,
            team_id=dao_id,
            user_id=f"tg:{user_id}",
            message=text,
            response=cached_answer,
            channel_id=chat_id,
            scope="short_term",
            save_agent_response=not is_service_response(cached_answer),
            agent_metadata={
                "cached_reply": True,
                "mentioned_bots": mentioned_bots,
                "requires_complex_reasoning": needs_complex_reasoning,
            },
            username=username,
        )
        return {"ok": True, "agent": agent_config.agent_id, "cached": True}
    
    # Check if there's a document context for follow-up questions
    session_id = f"telegram:{chat_id}"
    doc_context = await get_doc_context(session_id)
    
    # If there's a doc_id and the message looks like a question about the document
    if doc_context and doc_context.doc_id:
        # Check if it's a question (simple heuristic: contains question words or ends with ?)
        is_question = (
            "?" in text or
            any(word in text.lower() for word in ["що", "як", "чому", "коли", "де", "хто", "чи"])
        )
        
        if is_question:
            logger.info(f"{agent_config.name}: Follow-up question detected for doc_id={doc_context.doc_id}")
            # Try RAG query first
            rag_result = await ask_about_document(
                session_id=session_id,
                question=text,
                doc_id=doc_context.doc_id,
                dao_id=dao_id or doc_context.dao_id,
                user_id=f"tg:{user_id}"
            )
            
            if rag_result.success and rag_result.answer:
                # Truncate if too long for Telegram
                answer = rag_result.answer
                if len(answer) > TELEGRAM_SAFE_LENGTH:
                    answer = answer[:TELEGRAM_SAFE_LENGTH] + "\n\n_... (відповідь обрізано)_"
                
                await send_telegram_message(chat_id, answer, telegram_token)
                return {"ok": True, "agent": "parser", "mode": "rag_query"}
            # Fall through to regular chat if RAG query fails
    
    # ========================================
    # BEHAVIOR POLICY v2.1: Check if should respond
    # Gateway computes has_link and has_explicit_request (source of truth)
    # ========================================
    chat_type = chat.get("type", "private")
    is_private_chat = chat_type == "private"
    
    # Gateway: compute has_link (single source of truth)
    has_link = detect_url(text) if text else False
    
    # Gateway: detect mentioned agents
    mentioned_agents = []
    if text:
        for aid, variants in AGENT_NAME_VARIANTS.items():
            for v in variants:
                if v.lower() in text.lower():
                    mentioned_agents.append(aid)
                    break
    
    # Gateway: check if agent has been active in this chat recently (any user)
    agent_active_in_chat = has_agent_chat_participation(agent_config.agent_id, chat_id)
    
    # Gateway: compute has_explicit_request (single source of truth)
    # CONTRACT: imperative OR (? AND (dm OR reply OR mention OR thread))
    has_explicit_request = detect_explicit_request(
        text=text,
        is_dm=is_private_chat,
        is_reply_to_agent=is_reply_to_agent,
        mentioned_agents=mentioned_agents,
        thread_has_agent_participation=agent_active_in_chat,
    )
    
    # Check if this is a prober request (chat_id=0 or user_id=0)
    is_prober = is_prober_request(chat_id, user_id)
    
    # SOWA v2.2: 3-level decision (FULL / ACK / SILENT)
    sowa_decision = analyze_message(
        text=text,
        agent_id=agent_config.agent_id,
        chat_id=chat_id,
        user_id=str(user_id),
        has_media=has_link,
        media_caption=text if has_link else "",
        is_private_chat=is_private_chat,
        payload_explicit_request=has_explicit_request,
        payload_has_link=has_link,
        is_reply_to_agent=is_reply_to_agent,
        thread_has_agent_participation=agent_active_in_chat,
    )
    respond_decision = sowa_decision.should_respond
    respond_reason = sowa_decision.reason
    
    if sowa_decision.action == "SILENT":
        logger.info(f"\U0001f507 SOWA: Agent {agent_config.agent_id} NOT responding. Reason: {respond_reason}")
        # Save to memory for context tracking, but don't respond
        await memory_client.save_chat_turn(
            agent_id=agent_config.agent_id,
            team_id=dao_id,
            user_id=f"tg:{user_id}",
            message=text,
            response="",  # No response
            channel_id=chat_id,
            scope="short_term",
            save_agent_response=False,
            agent_metadata={
                "sowa_skipped": True,
                "skip_reason": respond_reason,
            },
            username=username,
        )
        return {"ok": True, "skipped": True, "reason": respond_reason}
    
    # ACK: send short presence message WITHOUT calling LLM/Router
    if sowa_decision.action == "ACK":
        ack_text = sowa_decision.ack_text or get_ack_text(agent_config.agent_id)
        logger.info(f"\U0001f44b SOWA ACK: Agent {agent_config.agent_id} sending ACK. Reason: {respond_reason}")
        
        # Send ACK to Telegram (no LLM call)
        if not is_prober:
            token = agent_config.get_telegram_token()
            if token:
                try:
                    url = f"https://api.telegram.org/bot{token}/sendMessage"
                    async with httpx.AsyncClient(timeout=30) as client:
                        ack_payload = {
                            "chat_id": chat_id,
                            "text": ack_text,
                        }
                        # Link ACK to the user's message for better UX
                        msg_id = update.message.get("message_id")
                        if msg_id:
                            ack_payload["reply_to_message_id"] = msg_id
                            ack_payload["allow_sending_without_reply"] = True
                        resp = await client.post(url, json=ack_payload)
                        if resp.status_code == 200:
                            logger.info(f"\U0001f44b ACK sent to chat {chat_id}: {ack_text}")
                        else:
                            logger.warning(f"ACK send failed: {resp.status_code} {resp.text[:200]}")
                except Exception as e:
                    logger.error(f"ACK send error: {e}")
        
        # Record ACK for cooldown and interaction tracking
        record_ack(agent_config.agent_id, str(chat_id))
        record_interaction(agent_config.agent_id, str(chat_id), str(user_id))
        
        # Save to memory
        await memory_client.save_chat_turn(
            agent_id=agent_config.agent_id,
            team_id=dao_id,
            user_id=f"tg:{user_id}",
            message=text,
            response=ack_text,
            channel_id=chat_id,
            scope="short_term",
            save_agent_response=True,
            agent_metadata={
                "sowa_ack": True,
                "ack_reason": respond_reason,
            },
            username=username,
        )
        return {"ok": True, "ack": True, "reason": respond_reason}
    
    # FULL: proceed with LLM/Router call
    # For prober requests, skip LLM/Router entirely to save tokens
    if is_prober:
        logger.info(f"\U0001f9ea PROBER: Agent {agent_config.agent_id} responding to prober (no LLM call). Reason: {respond_reason}")
        return {"ok": True, "agent": agent_config.agent_id, "prober": True, "response_preview": "[prober-skip-llm]"}
    else:
        logger.info(f"\u2705 SOWA: Agent {agent_config.agent_id} WILL respond (FULL). Reason: {respond_reason}")
    
    # Regular chat mode
    # Fetch memory context (includes local context as fallback)
    # Всі агенти мають доступ до однакової історії (80 повідомлень) для контексту
    context_limit = 80  # Однакове для всіх агентів
    memory_context = await memory_client.get_context(
        user_id=f"tg:{user_id}",
        agent_id=agent_config.agent_id,
        team_id=dao_id,
        channel_id=chat_id,
        limit=context_limit
    )
    
    # Build message with conversation context
    local_history = memory_context.get("local_context_text", "")
    
    # Check if this is a training group
    is_training_group = str(chat_id) in TRAINING_GROUP_IDS
    training_prefix = ""
    if is_training_group:
        training_prefix = "[РЕЖИМ НАВЧАННЯ - відповідай на це повідомлення, ти в навчальній групі Agent Preschool]\n\n"
        logger.info(f"🎓 Training mode activated for chat {chat_id}")
    
    if local_history:
        # Add conversation history to message for better context understanding
        message_with_context = f"{training_prefix}[Контекст розмови]\n{local_history}\n\n[Поточне повідомлення від {username}]\n{text}"
    else:
        message_with_context = f"{training_prefix}{text}"
    
    preferred_lang = await resolve_preferred_language_persistent(
        chat_id=chat_id,
        user_id=user_id,
        text=text or "",
        telegram_lang_code=from_user.get("language_code"),
        team_id=dao_id,
    )
    preferred_lang_label = preferred_language_label(preferred_lang)

    # Build request to Router
    system_prompt = agent_config.system_prompt
    logger.info(f"📝 {agent_config.name} system_prompt length: {len(system_prompt) if system_prompt else 0} chars")
    if system_prompt:
        logger.debug(f"System prompt preview: {system_prompt[:200]}...")
    else:
        logger.error(f"❌ {agent_config.name} system_prompt is EMPTY or None!")
    
    router_request = {
        "message": message_with_context,
        "mode": "chat",
        "agent": agent_config.agent_id,
        "metadata": {
            "source": "telegram",
            "dao_id": dao_id,
            "user_id": f"tg:{user_id}",
            "session_id": f"tg:{chat_id}:{dao_id}",
            "username": username,
            "chat_id": chat_id,
            "sender_is_bot": is_sender_bot,
            "mentioned_bots": mentioned_bots,
            "requires_complex_reasoning": needs_complex_reasoning,
            "is_reply_to_agent": is_reply_to_agent,
            "is_training_group": is_training_group,
            "preferred_response_language": preferred_lang,
            "preferred_response_language_label": preferred_lang_label,
        },
        "context": {
            "agent_name": agent_config.name,
            "system_prompt": system_prompt,
            "memory": memory_context,
            "participants": {
                "sender_is_bot": is_sender_bot,
                "mentioned_bots": mentioned_bots,
            },
        },
    }
    
    if should_force_detailed_reply(text):
        router_request["metadata"]["force_detailed"] = True

    if should_force_concise_reply(text):
        # IMPORTANT: preserve conversation context! Only append concise instruction
        router_request["metadata"]["force_concise"] = True
        router_request["message"] = (
            router_request["message"]
            + "\n\n(Інструкція: спочатку дай коротку відповідь по суті (1-3 абзаци), "
            "а якщо користувач попросить — розгорни детально.)"
            + f"\n(Мова відповіді: {preferred_lang_label}.)"
            + "\n(Не потрібно щоразу представлятися по імені або писати шаблонне: 'чим можу допомогти'.)"
        )
    
    # Helion policy: DeepSeek-first primary response path.
    if agent_config.agent_id == "helion":
        router_request["metadata"]["provider"] = "cloud_deepseek"
        router_request["metadata"]["reason"] = "helion_primary_deepseek"
    elif needs_complex_reasoning:
        router_request["metadata"]["provider"] = "cloud_deepseek"
        router_request["metadata"]["reason"] = "auto_complex"

    if not should_force_concise_reply(text):
        router_request["message"] = (
            router_request["message"]
            + f"\n\n(Мова відповіді: {preferred_lang_label}.)"
            + "\n(Не потрібно щоразу представлятися по імені або писати шаблонне: 'чим можу допомогти'.)"
        )
    
    # Send to Router
    logger.info(f"Sending to Router: agent={agent_config.agent_id}, dao={dao_id}, user=tg:{user_id}")
    response = await send_to_router(router_request)
    
    # Extract response
    if isinstance(response, dict) and response.get("ok"):
        answer_text = response.get("data", {}).get("text") or response.get("response", "")
        image_base64 = response.get("image_base64") or response.get("data", {}).get("image_base64")
        file_base64 = response.get("file_base64") or response.get("data", {}).get("file_base64")
        file_name = response.get("file_name") or response.get("data", {}).get("file_name") or "artifact.bin"
        file_mime = response.get("file_mime") or response.get("data", {}).get("file_mime") or "application/octet-stream"
        
        # Debug logging
        logger.info(f"📦 Router response: {len(answer_text)} chars, model={response.get('model')}, backend={response.get('backend')}")
        logger.info(f"📝 Response preview: {answer_text[:300]}..." if len(answer_text) > 300 else f"📝 Response: {answer_text}")
        if image_base64:
            logger.info(f"🖼️ Received image_base64: {len(image_base64)} chars")
        else:
            logger.debug("⚠️ No image_base64 in response")
        if file_base64:
            logger.info(f"📄 Received file_base64: {len(file_base64)} chars ({file_name})")

        # Check for NO_OUTPUT (LLM decided not to respond)
        if is_no_output_response(answer_text):
            logger.info(f"🔇 NO_OUTPUT: Agent {agent_config.agent_id} returned empty/NO_OUTPUT. Not sending to Telegram.")
            
            # P4: Detect NO_OUTPUT contract violations (extra text after NO_OUTPUT marker)
            _stripped = (answer_text or "").strip()
            _has_extra = False
            if _stripped and "__NO_OUTPUT__" in _stripped:
                _after_marker = _stripped.split("__NO_OUTPUT__", 1)[-1].strip()
                if _after_marker:
                    _has_extra = True
                    logger.warning(
                        f"🚨 policy_violation=no_output_extra_text "
                        f"agent={agent_config.agent_id} "
                        f"chat_id={chat_id} "
                        f"extra_text_len={len(_after_marker)} "
                        f"extra_preview={_after_marker[:80]!r}"
                    )
            elif _stripped and _stripped.lower() not in ("", "no_output", "no output", "silent", "мовчу", "—", ".", "..", "..."):
                # LLM returned something that looks like NO_OUTPUT but has unexpected content
                if len(_stripped) > 10:
                    _has_extra = True
                    logger.warning(
                        f"🚨 policy_violation=ambiguous_no_output "
                        f"agent={agent_config.agent_id} "
                        f"chat_id={chat_id} "
                        f"response_len={len(_stripped)} "
                        f"response_preview={_stripped[:80]!r}"
                    )
            
            # Save to memory for context tracking
            await memory_client.save_chat_turn(
                agent_id=agent_config.agent_id,
                team_id=dao_id,
                user_id=f"tg:{user_id}",
                message=text,
                response="",
                channel_id=chat_id,
                scope="short_term",
                save_agent_response=False,
                agent_metadata={
                    "no_output": True,
                    "original_response": answer_text[:100] if answer_text else "",
                    "policy_violation": "no_output_extra_text" if _has_extra else None,
                },
                username=username,
            )
            return {"ok": True, "skipped": True, "reason": "no_output_from_llm"}
        
        # Truncate if too long for Telegram
        if len(answer_text) > TELEGRAM_SAFE_LENGTH:
            answer_text = answer_text[:TELEGRAM_SAFE_LENGTH] + "\n\n_... (відповідь обрізано)_"
        
        # Skip Telegram sending for prober requests (chat_id=0)
        if is_prober:
            logger.info(f"🧪 PROBER: Skipping Telegram send for prober request. Response: {answer_text[:100]}...")
            return {"ok": True, "agent": agent_config.agent_id, "prober": True, "response_preview": answer_text[:100]}
        
        # Send file artifact if generated
        if file_base64:
            try:
                file_bytes = base64.b64decode(file_base64)
                token = telegram_token or os.getenv("TELEGRAM_BOT_TOKEN")
                url = f"https://api.telegram.org/bot{token}/sendDocument"
                caption = answer_text[:1024] if answer_text else ""
                safe_name = str(file_name).split("/")[-1].split("\\")[-1] or "artifact.bin"
                async with httpx.AsyncClient() as client:
                    files = {"document": (safe_name, BytesIO(file_bytes), file_mime)}
                    data = {"chat_id": chat_id}
                    if caption:
                        data["caption"] = caption
                    response_doc = await client.post(url, files=files, data=data, timeout=45.0)
                    response_doc.raise_for_status()
                    logger.info(f"✅ Sent generated document to Telegram chat {chat_id}: {safe_name}")
            except Exception as e:
                logger.error(f"❌ Failed to send document to Telegram: {e}")
                await send_telegram_message(chat_id, answer_text or "Файл згенеровано, але не вдалося надіслати документ.", telegram_token)
        # Send image if generated
        elif image_base64:
            try:
                # Decode base64 image
                image_bytes = base64.b64decode(image_base64)
                
                # Send photo to Telegram
                token = telegram_token or os.getenv("TELEGRAM_BOT_TOKEN")
                url = f"https://api.telegram.org/bot{token}/sendPhoto"
                
                async with httpx.AsyncClient() as client:
                    files = {"photo": ("image.png", BytesIO(image_bytes), "image/png")}
                    data = {"chat_id": chat_id, "caption": answer_text}
                    response_photo = await client.post(url, files=files, data=data, timeout=30.0)
                    response_photo.raise_for_status()
                    logger.info(f"✅ Sent generated image to Telegram chat {chat_id}")
            except Exception as e:
                logger.error(f"❌ Failed to send image to Telegram: {e}")
                # Fallback to text only
                await send_telegram_message(chat_id, answer_text, telegram_token)
        else:
            # Send text response only
            await send_telegram_message(chat_id, answer_text, telegram_token)
        
        # Record successful interaction for conversation context
        record_interaction(agent_config.agent_id, chat_id, str(user_id))
        
        await memory_client.save_chat_turn(
            agent_id=agent_config.agent_id,
            team_id=dao_id,
            user_id=f"tg:{user_id}",
            message=text,
            response=answer_text,
            channel_id=chat_id,
            scope="short_term",
            save_agent_response=not is_service_response(answer_text),
            agent_metadata={
                "mentioned_bots": mentioned_bots,
                "requires_complex_reasoning": needs_complex_reasoning,
                "preferred_language": preferred_lang,
            },
            username=username,
        )
        
        store_response_cache(agent_config.agent_id, chat_id, text, answer_text)
        
        return {"ok": True, "agent": agent_config.agent_id}
    else:
        error_msg = response.get("error", "Unknown error") if isinstance(response, dict) else "Router error"
        logger.error(f"Router error: {error_msg}")
        await send_telegram_message(chat_id, f"Вибач, сталася помилка: {error_msg}", telegram_token)
        return {"ok": False, "error": error_msg}


# ========================================
# Endpoints
# ========================================

# DAARWIZZ webhook endpoints (both paths for compatibility)

@router.get("/healthz")
async def healthz():
    try:
        from crews.agromatrix_crew.run import handle_message  # noqa: F401
        return {"ok": True, "status": "healthy"}
    except Exception as e:
        return {"ok": False, "status": "error", "error": str(e)}

@router.post("/telegram/webhook")
async def telegram_webhook(update: TelegramUpdate):
    """Handle Telegram webhook for DAARWIZZ agent (default path)."""
    try:
        return await handle_telegram_webhook(DAARWIZZ_CONFIG, update)
    except Exception as e:
        logger.error(f"Error handling DAARWIZZ Telegram webhook: {e}", exc_info=True)
        raise HTTPException(status_code=500, detail=str(e))


@router.post("/daarwizz/telegram/webhook")
async def daarwizz_telegram_webhook(update: TelegramUpdate):
    """Handle Telegram webhook for DAARWIZZ agent (agent-specific path)."""
    try:
        return await handle_telegram_webhook(DAARWIZZ_CONFIG, update)
    except Exception as e:
        logger.error(f"Error handling DAARWIZZ Telegram webhook: {e}", exc_info=True)
        raise HTTPException(status_code=500, detail=str(e))


# Legacy code - will be removed after testing
async def _old_telegram_webhook(update: TelegramUpdate):
    """Стара версія - використовується для тестування"""
    try:
        if not update.message:
            raise HTTPException(status_code=400, detail="No message in update")
        
        # Extract message details
        from_user = update.message.get("from", {})
        chat = update.message.get("chat", {})
        
        user_id = str(from_user.get("id", "unknown"))
        chat_id = str(chat.get("id", "unknown"))
        username = from_user.get("username", "")
        
        # Get DAO ID for this chat
        dao_id = get_dao_id(chat_id, "telegram", agent_id=agent_config.agent_id)
        
        # Check for /ingest command
        text = update.message.get("text", "")
        if text and text.strip().startswith("/ingest"):
            session_id = f"telegram:{chat_id}"
            
            # Check if there's a document in the message
            document = update.message.get("document")
            if document:
                mime_type = document.get("mime_type", "")
                file_name = document.get("file_name", "")
                file_id = document.get("file_id")
                
                is_pdf = (
                    mime_type == "application/pdf" or
                    (mime_type.startswith("application/") and file_name.lower().endswith(".pdf"))
                )
                
                if is_pdf and file_id:
                    try:
                        telegram_token = os.getenv("TELEGRAM_BOT_TOKEN")
                        file_path = await get_telegram_file_path(file_id)
                        if file_path:
                            file_url = f"https://api.telegram.org/file/bot{telegram_token}/{file_path}"
                            result = await ingest_document(
                                session_id=session_id,
                                doc_url=file_url,
                                file_name=file_name,
                                dao_id=dao_id,
                                user_id=f"tg:{user_id}"
                            )
                            
                            if result.success:
                                await send_telegram_message(
                                    chat_id,
                                    f"✅ **Документ імпортовано у RAG**\n\n"
                                    f"📊 Фрагментів: {result.ingested_chunks}\n"
                                    f"📁 DAO: {dao_id}\n\n"
                                    f"Тепер ти можеш задавати питання по цьому документу!"
                                )
                                return {"ok": True, "chunks_count": result.ingested_chunks}
                            else:
                                await send_telegram_message(chat_id, f"Вибач, не вдалося імпортувати: {result.error}")
                                return {"ok": False, "error": result.error}
                    except Exception as e:
                        logger.error(f"Ingest failed: {e}", exc_info=True)
                        await send_telegram_message(chat_id, "Вибач, не вдалося імпортувати документ.")
                        return {"ok": False, "error": "Ingest failed"}
            
            # Try to get last parsed doc_id from session context
            result = await ingest_document(
                session_id=session_id,
                dao_id=dao_id,
                user_id=f"tg:{user_id}"
            )
            
            if result.success:
                await send_telegram_message(
                    chat_id,
                    f"✅ **Документ імпортовано у RAG**\n\n"
                    f"📊 Фрагментів: {result.ingested_chunks}\n"
                    f"📁 DAO: {dao_id}\n\n"
                    f"Тепер ти можеш задавати питання по цьому документу!"
                )
                return {"ok": True, "chunks_count": result.ingested_chunks}
            else:
                await send_telegram_message(chat_id, "Спочатку надішли PDF-документ, а потім використай /ingest")
                return {"ok": False, "error": result.error}
        
        # Check if it's a document (PDF)
        document = update.message.get("document")
        if document:
            mime_type = document.get("mime_type", "")
            file_name = document.get("file_name", "")
            file_id = document.get("file_id")
            
            # Check if it's a PDF
            is_pdf = (
                mime_type == "application/pdf" or
                (mime_type.startswith("application/") and file_name.lower().endswith(".pdf"))
            )
            
            if is_pdf and file_id:
                logger.info(f"PDF document from {username} (tg:{user_id}), file_id: {file_id}, file_name: {file_name}")
                
                try:
                    # Get file path from Telegram
                    telegram_token = os.getenv("TELEGRAM_BOT_TOKEN")
                    file_path = await get_telegram_file_path(file_id)
                    if not file_path:
                        raise HTTPException(status_code=400, detail="Failed to get file from Telegram")
                    
                    # Build file URL
                    file_url = f"https://api.telegram.org/file/bot{telegram_token}/{file_path}"
                    
                    # Use doc_service for parsing
                    session_id = f"telegram:{chat_id}"
                    result = await parse_document(
                        session_id=session_id,
                        doc_url=file_url,
                        file_name=file_name,
                        dao_id=dao_id,
                        user_id=f"tg:{user_id}",
                        output_mode="qa_pairs",
                        metadata={"username": username, "chat_id": chat_id}
                    )
                    
                    if not result.success:
                        await send_telegram_message(chat_id, f"Вибач, не вдалося обробити документ: {result.error}")
                        return {"ok": False, "error": result.error}
                    
                    # Format response for Telegram
                    answer_text = ""
                    if result.qa_pairs:
                        # Convert QAItem to dict for formatting
                        qa_list = [{"question": qa.question, "answer": qa.answer} for qa in result.qa_pairs]
                        answer_text = format_qa_response(qa_list)
                    elif result.markdown:
                        answer_text = format_markdown_response(result.markdown)
                    elif result.chunks_meta and result.chunks_meta.get("chunks"):
                        chunks = result.chunks_meta.get("chunks", [])
                        answer_text = format_chunks_response(chunks)
                    else:
                        answer_text = "✅ Документ успішно оброблено, але формат відповіді не розпізнано."
                    
                    # Add hint about /ingest command
                    if not answer_text.endswith("_"):
                        answer_text += "\n\n💡 _Використай /ingest для імпорту документа у RAG_"
                    
                    logger.info(f"PDF parsing result: {len(answer_text)} chars, doc_id={result.doc_id}")
                    
                    # Send response back to Telegram
                    await send_telegram_message(chat_id, answer_text)
                    
                    return {"ok": True, "agent": "parser", "mode": "doc_parse", "doc_id": result.doc_id}
                    
                except Exception as e:
                    logger.error(f"PDF processing failed: {e}", exc_info=True)
                    await send_telegram_message(chat_id, "Вибач, не вдалося обробити PDF-документ. Переконайся, що файл не пошкоджений.")
                    return {"ok": False, "error": "PDF processing failed"}
            elif document and not is_pdf:
                # Non-PDF document
                await send_telegram_message(chat_id, "Наразі підтримуються тільки PDF-документи. Інші формати (docx, zip, тощо) будуть додані пізніше.")
                return {"ok": False, "error": "Unsupported document type"}
        
        # Check if it's a photo
        photo = update.message.get("photo")
        if photo:
            # Telegram sends multiple sizes, get the largest one (last in array)
            photo_obj = photo[-1] if isinstance(photo, list) else photo
            file_id = photo_obj.get("file_id") if isinstance(photo_obj, dict) else None
            
            if file_id:
                logger.info(f"Photo from {username} (tg:{user_id}), file_id: {file_id}")
                
                try:
                    # Get file path from Telegram
                    telegram_token = os.getenv("TELEGRAM_BOT_TOKEN")
                    file_path = await get_telegram_file_path(file_id)
                    if not file_path:
                        raise HTTPException(status_code=400, detail="Failed to get file from Telegram")
                    
                    # Build file URL
                    file_url = f"https://api.telegram.org/file/bot{telegram_token}/{file_path}"
                    
                    # Send to Router with specialist_vision_8b model (Swapper)
                    router_request = {
                        "message": f"Коротко (2-3 речення) опиши значення цього зображення: {file_url}",
                        "mode": "chat",
                        "agent": "daarwizz",
                        "metadata": {
                            "source": "telegram",
                            "dao_id": dao_id,
                            "user_id": f"tg:{user_id}",
                            "session_id": f"tg:{chat_id}:{dao_id}",
                            "username": username,
                            "chat_id": chat_id,
                            "file_id": file_id,
                            "file_url": file_url,
                            "has_image": True,
                        },
                        "context": {
                            "agent_name": DAARWIZZ_NAME,
                            "system_prompt": DAARWIZZ_SYSTEM_PROMPT,
                        },
                    }
                    
                    # Override LLM to use specialist_vision_8b for image understanding
                    router_request["metadata"]["use_llm"] = "specialist_vision_8b"
                    
                    # Send to Router
                    logger.info(f"Sending photo to Router with vision-8b: file_url={file_url[:50]}...")
                    response = await send_to_router(router_request)
                    
                    # Extract response
                    if isinstance(response, dict) and response.get("ok"):
                        answer_text = response.get("data", {}).get("text") or response.get("response", "")
                        
                        if answer_text:
                            # Photo processed successfully
                            await send_telegram_message(
                                chat_id,
                                answer_text  # No prefix, just the LLM response
                            )
                            
                            # Save to memory for context
                            await memory_client.save_chat_turn(
                                agent_id="daarwizz",
                                team_id=dao_id,
                                user_id=f"tg:{user_id}",
                                message=f"[Photo: {file_id}]",
                                response=answer_text,
                                channel_id=chat_id,
                                scope="short_term",
                                save_agent_response=not is_service_response(answer_text),
                                agent_metadata={"context": "photo"},
                                username=username,
                            )
                            
                            return {"ok": True, "agent": "daarwizz", "model": "specialist_vision_8b"}
                        else:
                            await send_telegram_message(chat_id, "Не вдалося отримати опис зображення.")
                            return {"ok": False, "error": "No description in response"}
                    else:
                        error_msg = response.get("error", "Unknown error") if isinstance(response, dict) else "Router error"
                        logger.error(f"Vision-8b error: {error_msg}")
                        await send_telegram_message(chat_id, f"Вибач, не вдалося обробити фото: {error_msg}")
                        return {"ok": False, "error": error_msg}
                    
                except Exception as e:
                    logger.error(f"Photo processing failed: {e}", exc_info=True)
                    await send_telegram_message(chat_id, "Вибач, сталася помилка при обробці фото.")
                    return {"ok": False, "error": "Photo processing failed"}
        
        # Check if it's a voice message
        voice = update.message.get("voice")
        audio = update.message.get("audio")
        video_note = update.message.get("video_note")
        
        text = ""
        
        if voice or audio or video_note:
            # Голосове повідомлення - розпізнаємо через STT
            media_obj = voice or audio or video_note
            file_id = media_obj.get("file_id") if media_obj else None
            
            if not file_id:
                raise HTTPException(status_code=400, detail="No file_id in voice/audio/video_note")
            
            logger.info(f"Voice message from {username} (tg:{user_id}), file_id: {file_id}")
            
            try:
                # Отримуємо файл з Telegram
                file_path = await get_telegram_file_path(file_id)
                if not file_path:
                    raise HTTPException(status_code=400, detail="Failed to get file from Telegram")
                
                # Завантажуємо файл
                file_url = f"https://api.telegram.org/file/bot{os.getenv('TELEGRAM_BOT_TOKEN')}/{file_path}"
                async with httpx.AsyncClient(timeout=30.0) as client:
                    file_resp = await client.get(file_url)
                    file_resp.raise_for_status()
                    audio_bytes = file_resp.content
                
                # Відправляємо в STT-сервіс
                stt_upload_url = _resolve_stt_upload_url()
                files = {"file": ("voice.ogg", audio_bytes, "audio/ogg")}
                
                async with httpx.AsyncClient(timeout=60.0) as client:
                    stt_resp = await client.post(stt_upload_url, files=files)
                    stt_resp.raise_for_status()
                    stt_data = stt_resp.json()
                    text = stt_data.get("text", "")
                
                logger.info(f"STT result: {text[:100]}...")
                
            except Exception as e:
                logger.error(f"STT processing failed: {e}", exc_info=True)
                await send_telegram_message(chat_id, "Вибач, не вдалося розпізнати голосове повідомлення. Спробуй надіслати текстом.", os.getenv("DAARWIZZ_TELEGRAM_BOT_TOKEN"))
                return {"ok": False, "error": "STT failed"}
        else:
            # Текстове повідомлення
            text = update.message.get("text", "")
            caption = update.message.get("caption", "")
            
            if not text and not caption:
                # Check for unsupported message types and silently ignore
                unsupported_types = ["sticker", "animation", "video_note", "contact", "location", 
                                    "venue", "poll", "dice", "game", "new_chat_members", 
                                    "left_chat_member", "new_chat_title", "new_chat_photo",
                                    "delete_chat_photo", "pinned_message"]
                for msg_type in unsupported_types:
                    if update.message.get(msg_type):
                        logger.debug(f"DAARWIZZ: Ignoring unsupported message type: {msg_type}")
                        return {"ok": True, "ignored": True, "reason": f"Unsupported message type: {msg_type}"}
                
                # If no supported content found, return silently
                return {"ok": True, "ignored": True, "reason": "No processable content"}
            
            # Use caption if text is empty
            if not text and caption:
                text = caption
        
        logger.info(f"Telegram message from {username} (tg:{user_id}) in chat {chat_id}: {text[:50]}")
        
        # Check if there's a document context for follow-up questions
        session_id = f"telegram:{chat_id}"
        doc_context = await get_doc_context(session_id)
        
        # If there's a doc_id and the message looks like a question about the document
        if doc_context and doc_context.doc_id:
            # Check if it's a question (simple heuristic: contains question words or ends with ?)
            is_question = (
                "?" in text or
                any(word in text.lower() for word in ["що", "як", "чому", "коли", "де", "хто", "чи"])
            )
            
            if is_question:
                logger.info(f"Follow-up question detected for doc_id={doc_context.doc_id}")
                # Try RAG query first
                rag_result = await ask_about_document(
                    session_id=session_id,
                    question=text,
                    doc_id=doc_context.doc_id,
                    dao_id=dao_id or doc_context.dao_id,
                    user_id=f"tg:{user_id}"
                )
                
                if rag_result.success and rag_result.answer:
                    # Truncate if too long for Telegram
                    answer = rag_result.answer
                    if len(answer) > TELEGRAM_SAFE_LENGTH:
                        answer = answer[:TELEGRAM_SAFE_LENGTH] + "\n\n_... (відповідь обрізано)_"
                    
                    await send_telegram_message(chat_id, answer)
                    return {"ok": True, "agent": "parser", "mode": "rag_query"}
                # Fall through to regular chat if RAG query fails
        
        # Regular chat mode
        # Fetch memory context
        memory_context = await memory_client.get_context(
            user_id=f"tg:{user_id}",
            agent_id="daarwizz",
            team_id=dao_id,
            channel_id=chat_id,
            limit=80
        )
        
        # Build request to Router with DAARWIZZ context
        router_request = {
            "message": text,
            "mode": "chat",
            "agent": "daarwizz",  # DAARWIZZ agent identifier
            "metadata": {
                "source": "telegram",
                "dao_id": dao_id,
                "user_id": f"tg:{user_id}",
                "session_id": f"tg:{chat_id}:{dao_id}",
                "username": username,
                "chat_id": chat_id,
            },
            "context": {
                "agent_name": DAARWIZZ_NAME,
                "system_prompt": DAARWIZZ_SYSTEM_PROMPT,
                "memory": memory_context,  # Додаємо пам'ять
                # RBAC context will be injected by Router
            },
        }
        
        # Send to Router
        logger.info(f"Sending to Router: agent=daarwizz, dao={dao_id}, user=tg:{user_id}")
        response = await send_to_router(router_request)
        
        # Extract response text
        if isinstance(response, dict):
            answer_text = response.get("data", {}).get("text") or response.get("response", "Вибач, я зараз не можу відповісти.")
        else:
            answer_text = "Вибач, сталася помилка."
        
        logger.info(f"Router response: {answer_text[:100]}")
        
        # Save chat turn to memory
        await memory_client.save_chat_turn(
            agent_id="daarwizz",
            team_id=dao_id,
            user_id=f"tg:{user_id}",
            message=text,
            response=answer_text,
            channel_id=chat_id,
            scope="short_term",
            save_agent_response=not is_service_response(answer_text),
            agent_metadata={"context": "legacy_daarwizz"},
            username=username,
        )
        
        # Send response back to Telegram
        await send_telegram_message(chat_id, answer_text, os.getenv("DAARWIZZ_TELEGRAM_BOT_TOKEN"))
        
        return {"ok": True, "agent": "daarwizz"}
        
    except Exception as e:
        logger.error(f"Error handling Telegram webhook: {e}", exc_info=True)
        raise HTTPException(status_code=500, detail=str(e))


@router.post("/discord/webhook")
async def discord_webhook(message: DiscordMessage):
    """
    Handle Discord webhook.
    
    Discord message format:
    {
      "content": "Hello!",
      "author": {"id": "123", "username": "alice"},
      "channel_id": "456",
      "guild_id": "789"
    }
    """
    try:
        if not message.content:
            raise HTTPException(status_code=400, detail="No content in message")
        
        # Extract message details
        text = message.content
        author = message.author or {}
        channel_id = message.channel_id or "unknown"
        guild_id = message.guild_id or "unknown"
        
        user_id = author.get("id", "unknown")
        username = author.get("username", "")
        
        # Get DAO ID for this channel
        dao_id = get_dao_id(channel_id, "discord")
        
        logger.info(f"Discord message from {username} (discord:{user_id}): {text[:50]}")
        
        # Fetch memory context
        memory_context = await memory_client.get_context(
            user_id=f"discord:{user_id}",
            agent_id="daarwizz",
            team_id=dao_id,
            channel_id=channel_id,
            limit=80
        )
        
        # Build request to Router with DAARWIZZ context
        router_request = {
            "message": text,
            "mode": "chat",
            "agent": "daarwizz",
            "metadata": {
                "source": "discord",
                "dao_id": dao_id,
                "user_id": f"discord:{user_id}",
                "session_id": f"discord:{channel_id}:{dao_id}",
                "username": username,
                "channel_id": channel_id,
                "guild_id": guild_id,
            },
            "context": {
                "agent_name": DAARWIZZ_NAME,
                "system_prompt": DAARWIZZ_SYSTEM_PROMPT,
                "memory": memory_context,  # Додаємо пам'ять
            },
        }
        
        # Send to Router
        response = await send_to_router(router_request)
        
        # Extract response text
        if isinstance(response, dict):
            answer_text = response.get("data", {}).get("text") or response.get("response", "Sorry, I can't respond right now.")
        else:
            answer_text = "Sorry, an error occurred."
        
        logger.info(f"Router response: {answer_text[:100]}")
        
        # Save chat turn to memory
        await memory_client.save_chat_turn(
            agent_id="daarwizz",
            team_id=dao_id,
            user_id=f"discord:{user_id}",
            message=text,
            response=answer_text,
            channel_id=channel_id,
            scope="short_term",
            save_agent_response=not is_service_response(answer_text),
            agent_metadata={"source": "discord"},
            username=username,
        )
        
        # TODO: Send response back to Discord
        # await send_discord_message(channel_id, answer_text)
        
        return {"ok": True, "agent": "daarwizz", "response": answer_text}
        
    except Exception as e:
        logger.error(f"Error handling Discord webhook: {e}", exc_info=True)
        raise HTTPException(status_code=500, detail=str(e))


# ========================================
# Helper Functions
# ========================================

async def get_telegram_file_path(file_id: str, bot_token: str = None) -> Optional[str]:
    """Отримати шлях до файлу з Telegram API"""
    telegram_token = bot_token or os.getenv("TELEGRAM_BOT_TOKEN")
    if not telegram_token:
        logger.error("TELEGRAM_BOT_TOKEN not set")
        return None
    
    url = f"https://api.telegram.org/bot{telegram_token}/getFile"
    params = {"file_id": file_id}
    
    try:
        async with httpx.AsyncClient(timeout=10.0) as client:
            response = await client.get(url, params=params)
            response.raise_for_status()
            data = response.json()
            if data.get("ok"):
                return data.get("result", {}).get("file_path")
    except Exception as e:
        logger.error(f"Error getting Telegram file: {e}")
    return None


def format_qa_response(qa_pairs: list, max_pairs: int = 5) -> str:
    """Format Q&A pairs for Telegram with length limits"""
    if not qa_pairs:
        return "📋 Документ оброблено, але Q&A пари не знайдено."
    
    qa_text = "📋 **Зміст документа:**\n\n"
    displayed = 0
    
    for i, qa in enumerate(qa_pairs[:max_pairs], 1):
        question = qa.get('question', 'Питання')
        answer = qa.get('answer', 'Відповідь')
        
        # Truncate answer if too long
        if len(answer) > 500:
            answer = answer[:500] + "..."
        
        pair_text = f"**{i}. {question}**\n{answer}\n\n"
        
        # Check if adding this pair would exceed limit
        if len(qa_text) + len(pair_text) > TELEGRAM_SAFE_LENGTH:
            break
        
        qa_text += pair_text
        displayed += 1
    
    if len(qa_pairs) > displayed:
        remaining = len(qa_pairs) - displayed
        qa_text += f"_... та ще {remaining} {'питань' if remaining > 1 else 'питання'}_"
    
    return qa_text


def format_markdown_response(markdown: str) -> str:
    """Format markdown response - returns raw text for LLM processing"""
    # Just return the text - LLM will summarize it
    return markdown


def format_chunks_response(chunks: list) -> str:
    """Format chunks summary for Telegram"""
    if not chunks:
        return "📄 Документ розпарсено, але фрагменти не знайдено."
    
    answer_text = f"📄 **Документ розпарсено** ({len(chunks)} фрагментів)\n\n"
    answer_text += "**Перші фрагменти:**\n\n"
    
    for i, chunk in enumerate(chunks[:3], 1):
        text = chunk.get('text', '')[:200]
        answer_text += f"{i}. {text}...\n\n"
    
    if len(chunks) > 3:
        answer_text += f"_... та ще {len(chunks) - 3} фрагментів_"
    
    return answer_text


def _zip_read_summary(markdown_text: str) -> Optional[str]:
    """Extract a short summary of processed/skipped files from ZIP markdown."""
    if not markdown_text:
        return None
    lines = [line.strip() for line in markdown_text.splitlines()]
    try:
        processed = []
        skipped = []
        idx = 0
        while idx < len(lines):
            line = lines[idx]
            if line.lower() == "processed files:":
                idx += 1
                while idx < len(lines) and lines[idx].startswith("- "):
                    processed.append(lines[idx][2:].strip())
                    idx += 1
                continue
            if line.lower() == "skipped files:":
                idx += 1
                while idx < len(lines) and lines[idx].startswith("- "):
                    skipped.append(lines[idx][2:].strip())
                    idx += 1
                continue
            idx += 1
        if not processed and not skipped:
            return None
        processed_text = ", ".join(processed) if processed else "нічого"
        skipped_text = ", ".join(skipped) if skipped else "нічого"
        return f"Прочитала з ZIP: {processed_text}; пропустила: {skipped_text}."
    except Exception:
        return None


def _default_theme_payload(brand_id: str) -> Dict[str, Any]:
    """Return a minimal theme.json payload for quick publish."""
    return {
        "theme_version": "v1.0.0",
        "brand_id": brand_id,
        "layout": {
            "page": "LAYOUT_WIDE",
            "safe_area": {"x": 0.6, "y": 0.45, "w": 12.1, "h": 6.2},
        },
        "palette": {
            "primary": "#0B1220",
            "secondary": "#1E293B",
            "accent": "#22C55E",
            "bg": "#FFFFFF",
            "text": "#0F172A",
        },
        "typography": {
            "font_primary": "Inter",
            "font_secondary": "Inter",
            "sizes": {"h1": 38, "h2": 28, "h3": 22, "body": 16, "small": 12},
            "weights": {"regular": 400, "medium": 500, "bold": 700},
            "line_height": {"tight": 1.05, "normal": 1.15, "relaxed": 1.25},
        },
        "components": {
            "header": {"enabled": True, "logo_variant": "light", "show_title": False},
            "footer": {"enabled": True, "show_page_number": True, "left_text": brand_id},
        },
        "rules": {
            "max_bullets": 6,
            "max_bullet_len": 110,
            "min_font_body": 12,
            "overflow_strategy": "appendix",
        },
    }


async def _brand_intake_request(payload: Dict[str, Any]) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.post(f"{BRAND_INTAKE_URL}/brand/intake", json=payload)
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Brand intake error: {resp.text[:200]}")
        return resp.json()


async def _brand_publish_theme(brand_id: str, theme: Dict[str, Any], theme_version: Optional[str]) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.post(
            f"{BRAND_REGISTRY_URL}/brands/{brand_id}/themes",
            json={"theme": theme, "theme_version": theme_version},
        )
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Brand publish error: {resp.text[:200]}")
        return resp.json()


async def _brand_get_latest(brand_id: str) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.get(f"{BRAND_REGISTRY_URL}/brands/{brand_id}/latest")
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Brand latest error: {resp.text[:200]}")
        return resp.json()


async def _brand_get_theme(brand_id: str, theme_version: str) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.get(f"{BRAND_REGISTRY_URL}/brands/{brand_id}/themes/{theme_version}")
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Brand get error: {resp.text[:200]}")
        return resp.json()


async def _presentation_render(slidespec: Dict[str, Any], brand_id: str, theme_version: str) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=30.0) as client:
        # Call presentation-renderer directly
        resp = await client.post(
            f"{PRESENTATION_RENDERER_URL}/present/render",
            json={
                "brand_id": brand_id,
                "theme_version": theme_version or "v1.0.0",
                "slidespec": slidespec,
                "output": "pptx"
            },
        )
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Presentation render error: {resp.text[:200]}")
        result = resp.json()
        # Return consistent response format
        return {
            "render_id": result.get("render_id"),
            "artifact_id": result.get("render_id"),  # Use render_id as artifact_id for now
            "job_id": result.get("render_id"),
            "status": result.get("status"),
            "brand_id": brand_id,
            "theme_version": theme_version
        }


def _sanitize_error_text(text: str) -> str:
    if not text:
        return ""
    sanitized = text
    sanitized = re.sub(r"https?://\\S+", "[url]", sanitized)
    sanitized = re.sub(r"[A-Za-z0-9_-]{20,}", "[token]", sanitized)
    return sanitized[:400]


def _can_access_artifact(acl_ref: Optional[str], dao_id: Optional[str], user_id: str) -> bool:
    if not acl_ref:
        return True
    acl = acl_ref.lower()
    if "public" in acl:
        return True
    if dao_id and str(dao_id).lower() in acl:
        return True
    if user_id and user_id.lower() in acl:
        return True
    return False


async def _artifact_get(artifact_id: str) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.get(f"{ARTIFACT_REGISTRY_URL}/artifacts/{artifact_id}")
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Artifact get error: {resp.text[:200]}")
        return resp.json()


async def _artifact_versions(artifact_id: str) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.get(f"{ARTIFACT_REGISTRY_URL}/artifacts/{artifact_id}/versions")
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Artifact versions error: {resp.text[:200]}")
        return resp.json()


async def _artifact_job_status(job_id: str) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.get(f"{ARTIFACT_REGISTRY_URL}/jobs/{job_id}")
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Job status error: {resp.text[:200]}")
        return resp.json()


def _format_job_status_message(job: Dict[str, Any], job_id: str) -> str:
    status = job.get("status")
    job_type = job.get("job_type")
    artifact_id = job.get("artifact_id")
    input_version_id = job.get("input_version_id")
    output_version_id = job.get("output_version_id")
    error_text = _sanitize_error_text(job.get("error_text", ""))
    meta = job.get("meta_json") or {}
    if isinstance(meta, str):
        try:
            meta = json.loads(meta)
        except Exception:
            meta = {}

    meta_bits = []
    for key in ["chunks_count", "parser_version", "chunker_version", "index_fingerprint", "fingerprint", "parsed_version_id", "chunks_version_id"]:
        if meta.get(key):
            value = meta.get(key)
            if key in {"fingerprint", "index_fingerprint"}:
                value = str(value)[:16] + "…"
            meta_bits.append(f"{key}: {value}")

    message = (
        "📌 **Статус job**\n\n"
        f"Status: `{status}`\n"
        f"Type: `{job_type}`\n"
        f"Job ID: `{job_id}`\n"
        f"Artifact ID: `{artifact_id}`\n"
        f"Input version: `{input_version_id}`"
    )
    if output_version_id:
        message += f"\nOutput version: `{output_version_id}`"
    if meta_bits:
        message += "\nMeta: " + "; ".join(meta_bits)
    if status == "failed" and error_text:
        message += f"\nПомилка: `{error_text}`"
    return message


async def _artifact_download(artifact_id: str, fmt: str) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.get(f"{ARTIFACT_REGISTRY_URL}/artifacts/{artifact_id}/download", params={"format": fmt})
        if resp.status_code >= 400:
            raise HTTPException(status_code=resp.status_code, detail=resp.text[:200])
        return resp.json()


async def _artifact_create_job(artifact_id: str, job_type: str, input_version_id: str) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.post(
            f"{ARTIFACT_REGISTRY_URL}/artifacts/{artifact_id}/jobs",
            json={"job_type": job_type, "input_version_id": input_version_id},
        )
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Create job error: {resp.text[:200]}")
        return resp.json()


async def _artifact_create(payload: Dict[str, Any]) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.post(f"{ARTIFACT_REGISTRY_URL}/artifacts", json=payload)
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Artifact create error: {resp.text[:200]}")
        return resp.json()


async def _artifact_add_version_from_url(artifact_id: str, payload: Dict[str, Any]) -> Dict[str, Any]:
    async with httpx.AsyncClient(timeout=60.0) as client:
        resp = await client.post(f"{ARTIFACT_REGISTRY_URL}/artifacts/{artifact_id}/versions/from_url", json=payload)
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Artifact version error: {resp.text[:200]}")
        return resp.json()


async def _artifact_job_done(job_id: str, note: str) -> None:
    async with httpx.AsyncClient(timeout=15.0) as client:
        resp = await client.post(f"{ARTIFACT_REGISTRY_URL}/jobs/{job_id}/done", json={"note": note})
        if resp.status_code >= 400:
            raise HTTPException(status_code=502, detail=f"Job done error: {resp.text[:200]}")


async def send_telegram_message(chat_id: str, text: str, bot_token: Optional[str] = None) -> bool:
    """Send message to Telegram chat with explicit error diagnostics."""
    telegram_token = bot_token or os.getenv("TELEGRAM_BOT_TOKEN")
    if not telegram_token:
        logger.error("TELEGRAM_BOT_TOKEN not set")
        return False

    # Defensive cleanup for occasional reasoning/markup leaks.
    import re
    safe_text = re.sub(r'<think>.*?</think>', '', text or "", flags=re.DOTALL)
    safe_text = re.sub(r'<think>.*$', '', safe_text, flags=re.DOTALL)
    safe_text = safe_text.strip() or "..."

    token_id = telegram_token.split(":", 1)[0] if ":" in telegram_token else "unknown"
    url = f"https://api.telegram.org/bot{telegram_token}/sendMessage"
    payload = {
        "chat_id": str(chat_id),
        "text": safe_text,
        "disable_web_page_preview": True,
    }

    try:
        async with httpx.AsyncClient() as client:
            response = await client.post(url, json=payload, timeout=15.0)

        if response.status_code >= 400:
            err_desc = response.text[:300]
            try:
                body = response.json()
                err_desc = body.get("description") or err_desc
            except Exception:
                pass
            logger.error(
                "Telegram sendMessage failed: bot_id=%s chat_id=%s status=%s desc=%s",
                token_id,
                chat_id,
                response.status_code,
                err_desc,
            )
            return False

        logger.info("Telegram message sent: bot_id=%s chat_id=%s", token_id, chat_id)
        return True
    except Exception as e:
        logger.error("Telegram sendMessage exception: bot_id=%s chat_id=%s error=%s", token_id, chat_id, e)
        return False


# ========================================
# Helion Telegram Webhook
# ========================================

@router.post("/helion/telegram/webhook")
async def helion_telegram_webhook(update: TelegramUpdate):
    """
    Handle Telegram webhook for Helion agent.
    """
    try:
        return await handle_telegram_webhook(HELION_CONFIG, update)
    except Exception as e:
        logger.error(f"Error handling Helion Telegram webhook: {e}", exc_info=True)
        raise HTTPException(status_code=500, detail=str(e))


# ========================================
# GREENFOOD Telegram Webhook
# ========================================

@router.post("/greenfood/telegram/webhook")
async def greenfood_telegram_webhook(update: TelegramUpdate):
    """
    Handle Telegram webhook for GREENFOOD agent.
    """
    try:
        return await handle_telegram_webhook(GREENFOOD_CONFIG, update)
    except Exception as e:
        logger.error(f"Error handling GREENFOOD Telegram webhook: {e}", exc_info=True)
        raise HTTPException(status_code=500, detail=str(e))


# ========================================
# NUTRA Telegram Webhook
# ========================================

@router.post("/nutra/telegram/webhook")
async def nutra_telegram_webhook(update: TelegramUpdate):
    """
    Handle Telegram webhook for NUTRA agent.
    """
    try:
        return await handle_telegram_webhook(NUTRA_CONFIG, update)
    except Exception as e:
        logger.error(f"Error handling NUTRA Telegram webhook: {e}", exc_info=True)
        raise HTTPException(status_code=500, detail=str(e))


# Legacy code - will be removed after testing
async def _old_helion_telegram_webhook(update: TelegramUpdate):
    """Стара версія - використовується для тестування"""
    try:
        if not update.message:
            raise HTTPException(status_code=400, detail="No message in update")
        
        # Extract message details
        from_user = update.message.get("from", {})
        chat = update.message.get("chat", {})
        
        user_id = str(from_user.get("id", "unknown"))
        chat_id = str(chat.get("id", "unknown"))
        username = from_user.get("username", "")
        
        # Get DAO ID for this chat (Energy Union specific)
        dao_id = get_dao_id(chat_id, "telegram", agent_id=agent_config.agent_id)
        
        # Check for /ingest command
        text = update.message.get("text", "")
        if text and text.strip().startswith("/ingest"):
            session_id = f"telegram:{chat_id}"
            
            # Check if there's a document in the message
            document = update.message.get("document")
            if document:
                mime_type = document.get("mime_type", "")
                file_name = document.get("file_name", "")
                file_id = document.get("file_id")
                
                is_pdf = (
                    mime_type == "application/pdf" or
                    (mime_type.startswith("application/") and file_name.lower().endswith(".pdf"))
                )
                
                if is_pdf and file_id:
                    try:
                        helion_token = os.getenv("HELION_TELEGRAM_BOT_TOKEN")
                        file_path = await get_telegram_file_path(file_id)
                        if file_path:
                            file_url = f"https://api.telegram.org/file/bot{helion_token}/{file_path}"
                            result = await ingest_document(
                                session_id=session_id,
                                doc_url=file_url,
                                file_name=file_name,
                                dao_id=dao_id,
                                user_id=f"tg:{user_id}"
                            )
                            
                            if result.success:
                                await send_telegram_message(
                                    chat_id,
                                    f"✅ **Документ імпортовано у RAG**\n\n"
                                    f"📊 Фрагментів: {result.ingested_chunks}\n"
                                    f"📁 DAO: {dao_id}\n\n"
                                    f"Тепер ти можеш задавати питання по цьому документу!",
                                    helion_token
                                )
                                return {"ok": True, "chunks_count": result.ingested_chunks}
                            else:
                                await send_telegram_message(chat_id, f"Вибач, не вдалося імпортувати: {result.error}", helion_token)
                                return {"ok": False, "error": result.error}
                    except Exception as e:
                        logger.error(f"Helion: Ingest failed: {e}", exc_info=True)
                        await send_telegram_message(chat_id, "Вибач, не вдалося імпортувати документ.", helion_token)
                        return {"ok": False, "error": "Ingest failed"}
            
            # Try to get last parsed doc_id from session context
            helion_token = os.getenv("HELION_TELEGRAM_BOT_TOKEN")
            result = await ingest_document(
                session_id=session_id,
                dao_id=dao_id,
                user_id=f"tg:{user_id}"
            )
            
            if result.success:
                await send_telegram_message(
                    chat_id,
                    f"✅ **Документ імпортовано у RAG**\n\n"
                    f"📊 Фрагментів: {result.ingested_chunks}\n"
                    f"📁 DAO: {dao_id}\n\n"
                    f"Тепер ти можеш задавати питання по цьому документу!",
                    helion_token
                )
                return {"ok": True, "chunks_count": result.ingested_chunks}
            else:
                await send_telegram_message(chat_id, "Спочатку надішли PDF-документ, а потім використай /ingest", helion_token)
                return {"ok": False, "error": result.error}
        
        # Check if it's a document (PDF)
        document = update.message.get("document")
        if document:
            mime_type = document.get("mime_type", "")
            file_name = document.get("file_name", "")
            file_id = document.get("file_id")
            
            is_pdf = (
                mime_type == "application/pdf" or
                (mime_type.startswith("application/") and file_name.lower().endswith(".pdf"))
            )
            
            if is_pdf and file_id:
                logger.info(f"Helion: PDF document from {username} (tg:{user_id}), file_id: {file_id}, file_name: {file_name}")
                
                try:
                    helion_token = os.getenv("HELION_TELEGRAM_BOT_TOKEN")
                    file_path = await get_telegram_file_path(file_id)
                    if not file_path:
                        raise HTTPException(status_code=400, detail="Failed to get file from Telegram")
                    
                    file_url = f"https://api.telegram.org/file/bot{helion_token}/{file_path}"
                    
                    session_id = f"telegram:{chat_id}"
                    result = await parse_document(
                        session_id=session_id,
                        doc_url=file_url,
                        file_name=file_name,
                        dao_id=dao_id,
                        user_id=f"tg:{user_id}",
                        output_mode="qa_pairs",
                        metadata={"username": username, "chat_id": chat_id}
                    )
                    
                    if not result.success:
                        await send_telegram_message(chat_id, f"Вибач, не вдалося обробити документ: {result.error}", helion_token)
                        return {"ok": False, "error": result.error}
                    
                    # Format response for Telegram
                    answer_text = ""
                    if result.qa_pairs:
                        qa_list = [{"question": qa.question, "answer": qa.answer} for qa in result.qa_pairs]
                        answer_text = format_qa_response(qa_list)
                    elif result.markdown:
                        answer_text = format_markdown_response(result.markdown)
                    elif result.chunks_meta and result.chunks_meta.get("chunks"):
                        chunks = result.chunks_meta.get("chunks", [])
                        answer_text = format_chunks_response(chunks)
                    else:
                        answer_text = "✅ Документ успішно оброблено, але формат відповіді не розпізнано."
                    
                    if not answer_text.endswith("_"):
                        answer_text += "\n\n💡 _Використай /ingest для імпорту документа у RAG_"
                    
                    logger.info(f"Helion: PDF parsing result: {len(answer_text)} chars, doc_id={result.doc_id}")
                    await send_telegram_message(chat_id, answer_text, helion_token)
                    return {"ok": True, "agent": "parser", "mode": "doc_parse", "doc_id": result.doc_id}
                    
                except Exception as e:
                    logger.error(f"Helion: PDF processing failed: {e}", exc_info=True)
                    await send_telegram_message(chat_id, "Вибач, не вдалося обробити PDF-документ. Переконайся, що файл не пошкоджений.", helion_token)
                    return {"ok": False, "error": "PDF processing failed"}
            elif document and not is_pdf:
                helion_token = os.getenv("HELION_TELEGRAM_BOT_TOKEN")
                await send_telegram_message(chat_id, "Наразі підтримуються тільки PDF-документи. Інші формати (docx, zip, тощо) будуть додані пізніше.", helion_token)
                return {"ok": False, "error": "Unsupported document type"}
        
        # Check if it's a photo
        photo = update.message.get("photo")
        if photo:
            # Telegram sends multiple sizes, get the largest one (last in array)
            photo_obj = photo[-1] if isinstance(photo, list) else photo
            file_id = photo_obj.get("file_id") if isinstance(photo_obj, dict) else None
            
            if file_id:
                logger.info(f"Helion: Photo from {username} (tg:{user_id}), file_id: {file_id}")
                
                try:
                    # Get file path from Telegram
                    helion_token = os.getenv("HELION_TELEGRAM_BOT_TOKEN")
                    file_path = await get_telegram_file_path(file_id, helion_token)
                    if not file_path:
                        raise HTTPException(status_code=400, detail="Failed to get file from Telegram")
                    
                    # Build file URL
                    file_url = f"https://api.telegram.org/file/bot{helion_token}/{file_path}"
                    
                    # Send to Router with specialist_vision_8b model (Swapper)
                    # IMPORTANT: Request BRIEF description (2-3 sentences per v2.3 prompt rules)
                    router_request = {
                        "message": f"Коротко (2-3 речення максимум): що на цьому зображенні та яке його значення для Energy Union? {file_url}",
                        "mode": "chat",
                        "agent": "helion",
                        "metadata": {
                            "source": "telegram",
                            "dao_id": dao_id,
                            "user_id": f"tg:{user_id}",
                            "session_id": f"tg:{chat_id}:{dao_id}",
                            "username": username,
                            "chat_id": chat_id,
                            "file_id": file_id,
                            "file_url": file_url,
                            "has_image": True,
                        },
                        "context": {
                            "agent_name": HELION_NAME,
                            "system_prompt": HELION_SYSTEM_PROMPT,
                        },
                    }
                    
                    # Override LLM to use specialist_vision_8b for image understanding
                    router_request["metadata"]["use_llm"] = "specialist_vision_8b"
                    
                    # Send to Router
                    logger.info(f"Helion: Sending photo to Router with vision-8b: file_url={file_url[:50]}...")
                    response = await send_to_router(router_request)
                    
                    # Extract response
                    if isinstance(response, dict) and response.get("ok"):
                        answer_text = response.get("data", {}).get("text") or response.get("response", "")
                        
                        if answer_text:
                            # Photo processed - send LLM response directly WITHOUT prefix
                            await send_telegram_message(
                                chat_id,
                                answer_text,  # No prefix, just the LLM response
                                helion_token
                            )
                            
                            # Save to memory for context
                            await memory_client.save_chat_turn(
                                agent_id="helion",
                                team_id=dao_id,
                                user_id=f"tg:{user_id}",
                                message=f"[Photo: {file_id}]",
                                response=answer_text,
                                channel_id=chat_id,
                                scope="short_term",
                                save_agent_response=not is_service_response(answer_text),
                                agent_metadata={"context": "photo"},
                                username=username,
                            )
                            
                            return {"ok": True, "agent": "helion", "model": "specialist_vision_8b"}
                        else:
                            await send_telegram_message(chat_id, "Не вдалося отримати опис зображення.", helion_token)
                            return {"ok": False, "error": "No description in response"}
                    else:
                        error_msg = response.get("error", "Unknown error") if isinstance(response, dict) else "Router error"
                        logger.error(f"Helion: Vision-8b error: {error_msg}")
                        await send_telegram_message(chat_id, "Вибач, сталася помилка при обробці фото.", helion_token)
                        return {"ok": False, "error": error_msg}
                    
                except Exception as e:
                    logger.error(f"Helion: Photo processing failed: {e}", exc_info=True)
                    helion_token = os.getenv("HELION_TELEGRAM_BOT_TOKEN")
                    await send_telegram_message(chat_id, "Вибач, сталася помилка при обробці фото.", helion_token)
                    return {"ok": False, "error": "Photo processing failed"}
        
        # Get message text
        text = update.message.get("text", "")
        if not text:
            raise HTTPException(status_code=400, detail="No text in message")
        
        logger.info(f"Helion Telegram message from {username} (tg:{user_id}) in chat {chat_id}: {text[:50]}")
        mentioned_bots = extract_bot_mentions(text)
        needs_complex_reasoning = requires_complex_reasoning(text)
        
        # Check if there's a document context for follow-up questions
        session_id = f"telegram:{chat_id}"
        doc_context = await get_doc_context(session_id)
        
        # If there's a doc_id and the message looks like a question about the document
        if doc_context and doc_context.doc_id:
            # Check if it's a question (simple heuristic: contains question words or ends with ?)
            is_question = (
                "?" in text or
                any(word in text.lower() for word in ["що", "як", "чому", "коли", "де", "хто", "чи"])
            )
            
            if is_question:
                logger.info(f"Helion: Follow-up question detected for doc_id={doc_context.doc_id}")
                # Try RAG query first
                rag_result = await ask_about_document(
                    session_id=session_id,
                    question=text,
                    doc_id=doc_context.doc_id,
                    dao_id=dao_id or doc_context.dao_id,
                    user_id=f"tg:{user_id}"
                )
                
                if rag_result.success and rag_result.answer:
                    # Truncate if too long for Telegram
                    answer = rag_result.answer
                    if len(answer) > TELEGRAM_SAFE_LENGTH:
                        answer = answer[:TELEGRAM_SAFE_LENGTH] + "\n\n_... (відповідь обрізано)_"
                    
                    helion_token = os.getenv("HELION_TELEGRAM_BOT_TOKEN")
                    await send_telegram_message(chat_id, answer, helion_token)
                    return {"ok": True, "agent": "parser", "mode": "rag_query"}
                # Fall through to regular chat if RAG query fails
        
        # Regular chat mode
        # Fetch memory context (includes local context as fallback)
        # All agents use limit=80 for full conversation history
        memory_context = await memory_client.get_context(
            user_id=f"tg:{user_id}",
            agent_id="helion",
            team_id=dao_id,
            channel_id=chat_id,
            limit=80
        )
        
        # Build message with conversation context
        local_history = memory_context.get("local_context_text", "")
        
        # Check if this is a training group
        is_training_group = str(chat_id) in TRAINING_GROUP_IDS
        training_prefix = ""
        if is_training_group:
            training_prefix = "[РЕЖИМ НАВЧАННЯ - відповідай на це повідомлення, ти в навчальній групі Agent Preschool]\n\n"
        
        if local_history:
            # Add conversation history to message for better context understanding
            message_with_context = f"{training_prefix}[Контекст розмови]\n{local_history}\n\n[Поточне повідомлення від {username}]\n{text}"
        else:
            message_with_context = f"{training_prefix}{text}"
        
        # Build request to Router with Helion context
        router_request = {
            "message": message_with_context,
            "mode": "chat",
            "agent": "helion",  # Helion agent identifier
            "metadata": {
                "source": "telegram",
                "dao_id": dao_id,
                "user_id": f"tg:{user_id}",
                "session_id": f"tg:{chat_id}:{dao_id}",
                "username": username,
                "chat_id": chat_id,
                "mentioned_bots": mentioned_bots,
                "requires_complex_reasoning": needs_complex_reasoning,
            },
            "context": {
                "agent_name": HELION_NAME,
                "system_prompt": HELION_SYSTEM_PROMPT,
                "memory": memory_context,
                # RBAC context will be injected by Router
            },
        }
        
        # Send to Router
        logger.info(f"Sending to Router: agent=helion, dao={dao_id}, user=tg:{user_id}")
        response = await send_to_router(router_request)
        
        # Extract response text
        if isinstance(response, dict):
            answer_text = response.get("data", {}).get("text") or response.get("response", "Вибач, я зараз не можу відповісти.")
        else:
            answer_text = "Вибач, сталася помилка."
        
        logger.info(f"Router response: {answer_text[:100]}")
        
        # Save chat turn to memory
        await memory_client.save_chat_turn(
            agent_id="helion",
            team_id=dao_id,
            user_id=f"tg:{user_id}",
            message=text,
            response=answer_text,
            channel_id=chat_id,
            scope="short_term",
            save_agent_response=not is_service_response(answer_text),
            agent_metadata={
                "context": "helion",
                "mentioned_bots": mentioned_bots,
                "requires_complex_reasoning": needs_complex_reasoning,
            },
            username=username,
        )
        
        # Send response back to Telegram
        await send_telegram_message(chat_id, answer_text, os.getenv("HELION_TELEGRAM_BOT_TOKEN"))
        
        return {"ok": True, "agent": "helion"}
        
    except Exception as e:
        logger.error(f"Error handling Helion Telegram webhook: {e}", exc_info=True)
        raise HTTPException(status_code=500, detail=str(e))


@router.get("/health")
async def health():
    """Health check endpoint"""
    agents_info = {}
    for agent_id, config in AGENT_REGISTRY.items():
        agents_info[agent_id] = {
            "name": config.name,
            "prompt_loaded": len(config.system_prompt) > 0,
            "telegram_token_configured": config.get_telegram_token() is not None
        }
    
    return {
        "status": "healthy",
        "agents": agents_info,
        "agents_count": len(AGENT_REGISTRY),
        "timestamp": datetime.utcnow().isoformat(),
    }


@router.post("/debug/agent_ping")
async def debug_agent_ping(request: dict = None):
    """
    E2E probe endpoint - tests full agent pipeline.
    Used by agent-e2e-prober for monitoring.
    Returns success only if router responds.
    """
    import time
    start = time.time()
    
    try:
        # Test 1: Check router connectivity
        router_url = os.getenv("ROUTER_URL", "http://router:8000")
        async with httpx.AsyncClient(timeout=10.0) as client:
            router_resp = await client.get(f"{router_url}/health")
            router_ok = router_resp.status_code == 200
        
        # Test 2: Check memory service connectivity
        memory_url = os.getenv("MEMORY_SERVICE_URL", "http://memory-service:8000")
        async with httpx.AsyncClient(timeout=10.0) as client:
            memory_resp = await client.get(f"{memory_url}/health")
            memory_ok = memory_resp.status_code == 200
        
        latency = time.time() - start
        
        return {
            "success": router_ok and memory_ok,
            "latency_seconds": round(latency, 3),
            "checks": {
                "router": router_ok,
                "memory_service": memory_ok,
            },
            "timestamp": datetime.utcnow().isoformat(),
        }
    except Exception as e:
        return {
            "success": False,
            "error": str(e)[:100],
            "latency_seconds": round(time.time() - start, 3),
            "timestamp": datetime.utcnow().isoformat(),
        }