graph.py

# graph.py - Fixed version with proper state handling for concurrent updates

import json
import re
import math
import os
import uuid
import shutil
import zipfile
import operator
from typing import TypedDict, List, Dict, Optional, Annotated, Any
from datetime import datetime
from langchain_openai import ChatOpenAI
from langgraph.graph import StateGraph, END
from memory_manager import memory_manager
from code_executor import execute_python_code
from logging_config import setup_logging, get_logger

# Artifact libs
import nbformat
from nbformat.v4 import new_notebook, new_markdown_cell, new_code_cell
import pandas as pd
from docx import Document
from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer
from reportlab.lib.styles import getSampleStyleSheet

# Add after other imports
from multi_language_support import (
    detect_language,
    extract_code_blocks_multi_lang,
    execute_code,
    detect_requested_output_types_enhanced,
    write_script_multi_lang,
    LANGUAGES
)

# Replace existing functions
detect_requested_output_types = detect_requested_output_types_enhanced
write_script = write_script_multi_lang

# --- Configuration ---
OUT_DIR = os.environ.get("OUT_DIR", "/tmp")
os.makedirs(OUT_DIR, exist_ok=True)
EXPORTS_DIR = os.path.join(OUT_DIR, "exports")
os.makedirs(EXPORTS_DIR, exist_ok=True)

# --- Helpers ---
def ensure_list(state, key):
    v = state.get(key) if state else None
    if v is None:
        return []
    if isinstance(v, list):
        return v
    if isinstance(v, tuple):
        return list(v)
    return [v]

def ensure_int(state, key, default=0):
    try:
        v = state.get(key) if state else None
        if v is None:
            return default
        return int(v)
    except Exception:
        return default

def sanitize_path(path: str) -> str:
    return os.path.abspath(path)

# --- Setup ---
setup_logging()
log = get_logger(__name__)
INITIAL_MAX_REWORK_CYCLES = 3
GPT4O_INPUT_COST_PER_1K_TOKENS = 0.005
GPT4O_OUTPUT_COST_PER_1K_TOKENS = 0.015
AVG_TOKENS_PER_CALL = 2.0

# --- State ---
class AgentState(TypedDict):
    userInput: str
    chatHistory: List[str]
    coreObjectivePrompt: str
    retrievedMemory: Optional[str]
    pmPlan: Dict
    experimentCode: Optional[str]
    experimentResults: Optional[Dict]
    draftResponse: str
    qaFeedback: Optional[str]
    approved: bool
    execution_path: Annotated[List[str], operator.add]
    rework_cycles: int
    max_loops: int
    # Use Annotated with operator.add for fields that multiple agents might update
    status_updates: Annotated[List[Dict[str, str]], operator.add]  # Changed from status_update
    current_cost: float
    budget_exceeded: bool
    # Add other fields that might have concurrent updates
    pragmatistReport: Optional[Dict]
    governanceReport: Optional[Dict]
    complianceReport: Optional[Dict]
    observerReport: Optional[Dict]
    knowledgeInsights: Optional[Dict]

# Helper to get latest status
def get_latest_status(state: AgentState) -> str:
    """Get the most recent status update from the list"""
    updates = state.get('status_updates', [])
    if updates and isinstance(updates, list):
        # Get the last update's status value
        for update in reversed(updates):
            if isinstance(update, dict) and 'status' in update:
                return update['status']
            elif isinstance(update, str):
                return update
    return "Processing..."

# Helper to add status update
def add_status_update(node_name: str, status: str) -> Dict[str, Any]:
    """Create a status update entry"""
    return {
        "status_updates": [{"node": node_name, "status": status, "timestamp": datetime.utcnow().isoformat()}]
    }

# --- LLM ---
llm = ChatOpenAI(model="gpt-4o", temperature=0.5, max_retries=3, request_timeout=60)

def parse_json_from_llm(llm_output: str) -> Optional[dict]:
    """
    More robust JSON extraction:
      - Looks for explicit ```json {} ``` blocks
      - Falls back to the last balanced {...} substring in the output
      - Tries ast.literal_eval for Python-like dicts
      - Performs conservative cleanup (remove trailing commas, comments, safe single->double quote) and retries
    Returns dict or None. Logs failures for debugging.
    """
    import re
    import json
    import ast
    from logging import getLogger

    logger = getLogger(__name__)

    if not llm_output or not isinstance(llm_output, str) or not llm_output.strip():
        return None

    text = llm_output.strip()

    # 1) explicit fenced JSON block
    match = re.search(r"```json\s*({.*?})\s*```", text, re.DOTALL | re.IGNORECASE)
    if match:
        candidate = match.group(1).strip()
        try:
            return json.loads(candidate)
        except Exception as e:
            logger.debug(f"json.loads failed on triple-backtick json block: {e}")

    # 2) any code-fence containing a JSON-like object
    match2 = re.search(r"```(?:json|python|text)?\s*({.*?})\s*```", text, re.DOTALL | re.IGNORECASE)
    if match2:
        candidate = match2.group(1).strip()
        try:
            return json.loads(candidate)
        except Exception as e:
            logger.debug(f"json.loads failed on fenced candidate: {e}")

    # 3) find first balanced {...} substring
    def find_balanced_brace_substring(s: str):
        start_idx = None
        depth = 0
        for i, ch in enumerate(s):
            if ch == '{':
                if start_idx is None:
                    start_idx = i
                depth += 1
            elif ch == '}':
                if depth > 0:
                    depth -= 1
                if depth == 0 and start_idx is not None:
                    return s[start_idx:i+1]
        return None

    candidate = find_balanced_brace_substring(text)

    # 4) fallback: last { ... } block heuristically
    if not candidate:
        first = text.find('{')
        last = text.rfind('}')
        if first != -1 and last != -1 and last > first:
            candidate = text[first:last+1]

    if candidate:
        # try json.loads directly
        try:
            return json.loads(candidate)
        except Exception as e:
            logger.debug(f"json.loads failed on candidate substring: {e}")

        # try ast.literal_eval (handles single quotes & Python literals)
        try:
            parsed = ast.literal_eval(candidate)
            if isinstance(parsed, (dict, list)):
                # convert to a strict JSON-compatible dict/list
                return json.loads(json.dumps(parsed))
        except Exception as e:
            logger.debug(f"ast.literal_eval failed: {e}")

        # conservative cleanup: remove comments, trailing commas, and handle simple single-quote strings
        cleaned = candidate
        try:
            # remove line comments //...
            cleaned = re.sub(r"//.*?$", "", cleaned, flags=re.MULTILINE)
            # remove block comments /* ... */
            cleaned = re.sub(r"/\*.*?\*/", "", cleaned, flags=re.DOTALL)
            # remove trailing commas before } or ]
            cleaned = re.sub(r",\s*([}\]])", r"\1", cleaned)

            # replace single-quoted strings with double quotes when likely safe
            def _single_to_double(m):
                inner = m.group(1)
                inner_escaped = inner.replace('"', '\\"')
                return f'"{inner_escaped}"'
            cleaned = re.sub(r"(?<=[:\{\[,]\s*)'([^']*?)'", _single_to_double, cleaned)

            # final attempt
            return json.loads(cleaned)
        except Exception as e:
            logger.debug(f"json.loads still failed after cleanup: {e}")

    # nothing parsed – log preview and return None
    logger.error("parse_json_from_llm failed to parse LLM output. LLM output preview (200 chars): %s", text[:200].replace("\n","\\n"))
    return None


# --- Artifact detection ---
KNOWN_ARTIFACT_TYPES = {"notebook","excel","word","pdf","image","repo","script"}

#def detect_requested_output_types(text: str) -> Dict:
#    if not text:
#        return {"requires_artifact": False, "artifact_type": None, "artifact_hint": None}
#    t = text.lower()
#    if any(k in t for k in ["jupyter notebook", "jupyter", "notebook", "ipynb"]):
#        return {"requires_artifact": True, "artifact_type": "notebook", "artifact_hint": "jupyter notebook"}
#    if any(k in t for k in ["excel", ".xlsx", "spreadsheet", "csv"]):
#        return {"requires_artifact": True, "artifact_type": "excel", "artifact_hint": "Excel file"}
#    if any(k in t for k in ["word document", ".docx", "docx"]):
#        return {"requires_artifact": True, "artifact_type": "word", "artifact_hint": "Word document"}
#    if any(k in t for k in ["pdf", "pdf file"]):
#        return {"requires_artifact": True, "artifact_type": "pdf", "artifact_hint": "PDF document"}
#    if any(k in t for k in ["repo", "repository", "app repo", "backend", "codebase"]):
#        return {"requires_artifact": True, "artifact_type": "repo", "artifact_hint": "application repository"}
#    if any(k in t for k in [".py", "python script", "script"]):
#        return {"requires_artifact": True, "artifact_type": "script", "artifact_hint": "Python script"}
#    return {"requires_artifact": False, "artifact_type": None, "artifact_hint": None}

def normalize_experiment_type(exp_type: Optional[str], goal_text: str) -> str:
    if not exp_type:
        detection = detect_requested_output_types(goal_text or "")
        return detection.get("artifact_type") or "word"
    s = exp_type.strip().lower()
    if s in KNOWN_ARTIFACT_TYPES:
        return s
    if "notebook" in s or "ipynb" in s:
        return "notebook"
    if "excel" in s or "xlsx" in s:
        return "excel"
    if "word" in s or "docx" in s:
        return "word"
    if "pdf" in s:
        return "pdf"
    if "repo" in s or "repository" in s or "backend" in s:
        return "repo"
    if "script" in s or "python" in s:
        return "script"
    detection = detect_requested_output_types(goal_text or "")
    return detection.get("artifact_type") or "word"

# --- Artifact builders ---
def write_notebook_from_text(llm_text: str, out_dir: Optional[str]=None) -> str:
    out_dir = out_dir or OUT_DIR
    os.makedirs(out_dir, exist_ok=True)
    code_blocks = re.findall(r"```python\s*(.*?)\s*```", llm_text, re.DOTALL)
    if not code_blocks:
        code_blocks = re.findall(r"```\s*(.*?)\s*```", llm_text, re.DOTALL)
    md_parts = re.split(r"```(?:python)?\s*.*?\s*```", llm_text, flags=re.DOTALL)
    nb = new_notebook()
    cells = []
    max_len = max(len(md_parts), len(code_blocks))
    for i in range(max_len):
        if i < len(md_parts) and md_parts[i].strip():
            cells.append(new_markdown_cell(md_parts[i].strip()))
        if i < len(code_blocks) and code_blocks[i].strip():
            cells.append(new_code_cell(code_blocks[i].strip()))
    if not cells:
        cells = [new_markdown_cell("# Notebook\n\nNo content generated.")]
    nb['cells'] = cells
    uid = uuid.uuid4().hex[:10]
    filename = os.path.join(out_dir, f"generated_notebook_{uid}.ipynb")
    nbformat.write(nb, filename)
    return filename

#def write_script(code_text: str, language_hint: Optional[str]=None, out_dir: Optional[str]=None) -> str:
#    out_dir = out_dir or OUT_DIR
#    os.makedirs(out_dir, exist_ok=True)
#    ext = ".txt"
#    if language_hint:
#        l = language_hint.lower()
#        if "python" in l:
#            ext = ".py"
#        elif "r" in l:
#            ext = ".R"
#        elif "java" in l:
#            ext = ".java"
#        elif "javascript" in l:
#            ext = ".js"
#    uid = uuid.uuid4().hex[:10]
#    filename = os.path.join(out_dir, f"generated_script_{uid}{ext}")
#    with open(filename, "w", encoding="utf-8") as f:
#        f.write(code_text)
#    return filename

def write_docx_from_text(text: str, out_dir: Optional[str]=None) -> str:
    out_dir = out_dir or OUT_DIR
    os.makedirs(out_dir, exist_ok=True)
    doc = Document()
    for para in [p.strip() for p in text.split("\n\n") if p.strip()]:
        doc.add_paragraph(para)
    uid = uuid.uuid4().hex[:10]
    filename = os.path.join(out_dir, f"generated_doc_{uid}.docx")
    doc.save(filename)
    return filename

def write_excel_from_tables(maybe_table_text: str, out_dir: Optional[str]=None) -> str:
    out_dir = out_dir or OUT_DIR
    os.makedirs(out_dir, exist_ok=True)
    uid = uuid.uuid4().hex[:10]
    filename = os.path.join(out_dir, f"generated_excel_{uid}.xlsx")
    try:
        try:
            parsed = json.loads(maybe_table_text)
            if isinstance(parsed, list):
                df = pd.DataFrame(parsed)
            elif isinstance(parsed, dict):
                df = pd.DataFrame([parsed])
            else:
                df = pd.DataFrame({"content":[str(maybe_table_text)]})
        except Exception:
            if "," in maybe_table_text:
                from io import StringIO
                df = pd.read_csv(StringIO(maybe_table_text))
            else:
                df = pd.DataFrame({"content":[maybe_table_text]})
        df.to_excel(filename, index=False, engine="openpyxl")
        return filename
    except Exception as e:
        log.error(f"Excel creation failed: {e}")
        return write_docx_from_text(f"Excel error: {e}\n\n{maybe_table_text}", out_dir=out_dir)

def write_pdf_from_text(text: str, out_dir: Optional[str]=None) -> str:
    out_dir = out_dir or OUT_DIR
    os.makedirs(out_dir, exist_ok=True)
    uid = uuid.uuid4().hex[:10]
    filename = os.path.join(out_dir, f"generated_doc_{uid}.pdf")
    try:
        doc = SimpleDocTemplate(filename)
        styles = getSampleStyleSheet()
        flowables = []
        for para in [p.strip() for p in text.split("\n\n") if p.strip()]:
            flowables.append(Paragraph(para.replace("\n","<br/>"), styles["Normal"]))
            flowables.append(Spacer(1, 8))
        doc.build(flowables)
        return filename
    except Exception as e:
        log.error(f"PDF creation failed: {e}")
        return write_docx_from_text(f"PDF error: {e}\n\n{text}", out_dir=out_dir)

def build_repo_zip(files_map: Dict[str,str], repo_name: str="generated_app", out_dir: Optional[str]=None) -> str:
    out_dir = out_dir or OUT_DIR
    os.makedirs(out_dir, exist_ok=True)
    uid = uuid.uuid4().hex[:8]
    repo_dir = os.path.join(out_dir, f"{repo_name}_{uid}")
    os.makedirs(repo_dir, exist_ok=True)
    
    for rel_path, content in files_map.items():
        dest = os.path.join(repo_dir, rel_path)
        os.makedirs(os.path.dirname(dest), exist_ok=True)
        
        if isinstance(content, str) and os.path.exists(content):
            shutil.copyfile(content, dest)
        else:
            with open(dest, "w", encoding="utf-8") as fh:
                fh.write(str(content))
    
    zip_path = os.path.join(out_dir, f"{repo_name}_{uid}.zip")
    with zipfile.ZipFile(zip_path, "w", zipfile.ZIP_DEFLATED) as zf:
        for root, _, files in os.walk(repo_dir):
            for f in files:
                full = os.path.join(root, f)
                arc = os.path.relpath(full, repo_dir)
                zf.write(full, arc)
    
    return zip_path

# --- Nodes ---
def run_triage_agent(state: AgentState):
    log.info("--- TRIAGE ---")
    prompt = f"Is this a greeting or a task? '{state.get('userInput','')}' Reply: 'greeting' or 'task'"
    response = llm.invoke(prompt)
    content = getattr(response, "content", "") or ""
    if 'greeting' in content.lower():
        return {
            "draftResponse": "Hello! How can I help?", 
            "execution_path": ["Triage"],
            **add_status_update("Triage", "Greeting")
        }
    return {
        "execution_path": ["Triage"],
        **add_status_update("Triage", "Task detected")
    }

def run_planner_agent(state: AgentState):
    log.info("--- PLANNER ---")
    path = ensure_list(state, 'execution_path') + ["Planner"]
    prompt = f"Create a plan for: '{state.get('userInput','')}'. JSON with 'plan' (list), 'estimated_llm_calls_per_loop' (int)"
    response = llm.invoke(prompt)
    plan_data = parse_json_from_llm(getattr(response, "content", "") or "")
    if not plan_data:
        return {
            "pmPlan": {"error": "Planning failed"}, 
            "execution_path": path,
            **add_status_update("Planner", "Error")
        }
    
    calls = plan_data.get('estimated_llm_calls_per_loop', 3)
    cost_per_loop = (calls * AVG_TOKENS_PER_CALL) * ((GPT4O_INPUT_COST_PER_1K_TOKENS + GPT4O_OUTPUT_COST_PER_1K_TOKENS) / 2)
    plan_data['max_loops_initial'] = INITIAL_MAX_REWORK_CYCLES
    plan_data['estimated_cost_usd'] = round(cost_per_loop * (INITIAL_MAX_REWORK_CYCLES + 1), 2)
    plan_data['cost_per_loop_usd'] = max(0.01, round(cost_per_loop, 3))
    
    detection = detect_requested_output_types(state.get('userInput',''))
    if detection.get('requires_artifact'):
        plan_data.setdefault('experiment_needed', True)
        plan_data.setdefault('experiment_type', detection.get('artifact_type'))
        plan_data.setdefault('experiment_goal', state.get('userInput',''))
    
    return {
        "pmPlan": plan_data, 
        "execution_path": path,
        **add_status_update("Planner", "Plan created")
    }

def run_memory_retrieval(state: AgentState):
    log.info("--- MEMORY ---")
    path = ensure_list(state, 'execution_path') + ["Memory"]
    mems = memory_manager.retrieve_relevant_memories(state.get('userInput',''))
    context = "\n".join([f"Memory: {m.page_content}" for m in mems]) if mems else "No memories"
    return {
        "retrievedMemory": context, 
        "execution_path": path,
        **add_status_update("Memory", "Memory retrieved")
    }

def run_intent_agent(state: AgentState):
    log.info("--- INTENT ---")
    path = ensure_list(state, 'execution_path') + ["Intent"]
    prompt = f"Refine into clear objective.\n\nMemory: {state.get('retrievedMemory')}\n\nRequest: {state.get('userInput','')}\n\nCore Objective:"
    response = llm.invoke(prompt)
    core_obj = getattr(response, "content", "") or ""
    return {
        "coreObjectivePrompt": core_obj, 
        "execution_path": path,
        **add_status_update("Intent", "Objective clarified")
    }

def run_pm_agent(state: AgentState):
    log.info("--- PM ---")
    # Ensure keys
    current_rework = ensure_int(state, 'rework_cycles', 0)
    max_loops_val = ensure_int(state, 'max_loops', INITIAL_MAX_REWORK_CYCLES)

    # If we've exhausted loops, short-circuit and produce fallback plan with a note
    if current_rework > max_loops_val:
        path = ensure_list(state, 'execution_path') + ["PM"]
        fallback_plan = {
            "plan_steps": ["Rework limit exceeded. Manual review required."],
            "experiment_needed": False,
            "experiment_type": "word",
            "experiment_goal": state.get('coreObjectivePrompt', state.get('userInput',''))
        }
        return {
            "pmPlan": fallback_plan, 
            "execution_path": path, 
            "rework_cycles": current_rework,
            **add_status_update("PM", "Rework limit hit - manual review")
        }

    # Normal behavior: increment rework count for this pass
    current_cycles = current_rework + 1
    path = ensure_list(state, 'execution_path') + ["PM"]

    context_parts = [
        f"=== USER REQUEST ===\n{state.get('userInput', '')}",
        f"\n=== OBJECTIVE ===\n{state.get('coreObjectivePrompt', '')}",
        f"\n=== MEMORY ===\n{state.get('retrievedMemory', 'None')}",
    ]
    if state.get('qaFeedback'):
        context_parts.append(f"\n=== QA FEEDBACK (MUST FIX) ===\n{state.get('qaFeedback')}")
        context_parts.append(f"\n=== PREVIOUS PLAN ===\n{json.dumps(state.get('pmPlan', {}).get('plan_steps', []), indent=2)}")
    full_context = "\n".join(context_parts)

    # Detect language preference
    detected = detect_requested_output_types_enhanced(state.get('userInput', ''))
    language_hint = LANGUAGES[detected['language']].name if detected.get('language') else "appropriate language"

    prompt = f"""Create DETAILED, EXECUTABLE plan.

{full_context}

If code generation is needed, use {language_hint}.

Return JSON with:
{{
    "plan_steps": [...],
    "experiment_needed": true/false,
    "experiment_type": "notebook|script|excel|word|pdf|repo",
    "experiment_goal": "...",
    "key_requirements": [...]
}}
Be concrete.
"""
    try:
        response = llm.invoke(prompt)
        plan = parse_json_from_llm(getattr(response, "content", "") or "")
    except Exception as e:
        log.warning("PM LLM failed: %s", e)
        plan = None

    if not plan:
        detection = detect_requested_output_types(state.get('userInput', ''))
        plan = {
            "plan_steps": ["Analyze request", "Process information", "Create deliverable", "Review"],
            "experiment_needed": detection.get('requires_artifact', False),
            "experiment_type": detection.get('artifact_type', 'word'),
            "experiment_goal": state.get('coreObjectivePrompt', state.get('userInput', '')),
            "key_requirements": []
        }

    # Normalize experiment_type
    exp_type = normalize_experiment_type(plan.get('experiment_type'), plan.get('experiment_goal',''))
    plan['experiment_type'] = exp_type
    if plan.get('experiment_needed') and not plan.get('experiment_goal'):
        plan['experiment_goal'] = state.get('userInput','')

    # Attach loop control info
    plan['max_loops_initial'] = max_loops_val
    plan['estimated_cost_usd'] = plan.get('estimated_cost_usd', 0.0)
    return {
        "pmPlan": plan, 
        "execution_path": path, 
        "rework_cycles": current_cycles, 
        "max_loops": max_loops_val,
        **add_status_update("PM", f"Plan created ({len(plan.get('plan_steps', []))} steps)")
    }

def _extract_code_blocks(text: str, lang_hint: Optional[str]=None) -> List[str]:
    if lang_hint and "python" in (lang_hint or "").lower():
        blocks = re.findall(r"```python\s*(.*?)\s*```", text, re.DOTALL)
        if blocks:
            return blocks
    return re.findall(r"```(?:\w+)?\s*(.*?)\s*```", text, re.DOTALL)

def run_experimenter_agent(state: AgentState):
    log.info("--- EXPERIMENTER ---")
    path = ensure_list(state, 'execution_path') + ["Experimenter"]
    pm = state.get('pmPlan', {}) or {}
    
    if not pm.get('experiment_needed'):
        return {
            "experimentCode": None, 
            "experimentResults": None, 
            "execution_path": path,
            **add_status_update("Experimenter", "No experiment needed")
        }
    
    # Detect language from plan or request
    #detected = detect_requested_output_types_enhanced(pm.get('experiment_goal', ''))
    #language = detected.get('language', 'python')
    exp_type = normalize_experiment_type(pm.get('experiment_type'), pm.get('experiment_goal',''))
    goal = pm.get('experiment_goal', 'No goal')
    
    # BUILD RICH CONTEXT (this was missing!)
    context_parts = [
        f"=== USER REQUEST ===\n{state.get('userInput', '')}",
        f"\n=== OBJECTIVE ===\n{state.get('coreObjectivePrompt', '')}",
        f"\n=== PLAN ===\n{json.dumps(pm.get('plan_steps', []), indent=2)}",
        f"\n=== REQUIREMENTS ===\n{json.dumps(pm.get('key_requirements', []), indent=2)}",
    ]
    
    if state.get('retrievedMemory'):
        context_parts.append(f"\n=== CONTEXT ===\n{state.get('retrievedMemory', '')}")
    
    if state.get('qaFeedback'):
        context_parts.append(f"\n=== FEEDBACK TO ADDRESS ===\n{state.get('qaFeedback', '')}")
    
    full_context = "\n".join(context_parts)  # This line was missing!
    
    
    # Get language config
    from multi_language_support import LANGUAGES
    lang_config = LANGUAGES.get(language)
    lang_name = lang_config.name if lang_config else "Code"
    
    # Enhanced prompt with language specification
    enhanced_prompt = f"""Create HIGH-QUALITY {lang_name} {exp_type} artifact.

{full_context}

GOAL: {goal}

LANGUAGE: {lang_name}
REQUIREMENTS:
- Write idiomatic {lang_name} code following best practices
- Include appropriate comments and documentation
- Use language-specific features and libraries
- PRODUCTION-READY, COMPLETE content (NO placeholders)
- Include error handling appropriate for {lang_name}

Generate complete {lang_name} code:"""

    response = llm.invoke(enhanced_prompt)
    llm_text = getattr(response, "content", "") or ""
    
    # Extract code blocks with language detection
    from multi_language_support import extract_code_blocks_multi_lang
    code_blocks = extract_code_blocks_multi_lang(llm_text)
    
    if code_blocks:
        # Use the first detected language/code pair
        detected_lang, code_text = code_blocks[0]
        
        # Write script with proper extension
        script_path = write_script_multi_lang(code_text, detected_lang, out_dir=OUT_DIR)
        
        # Execute with appropriate runner
        exec_results = execute_code(code_text, detected_lang)
        
        results = {
            "success": exec_results.get("exit_code", 0) == 0,
            "paths": {"script": sanitize_path(script_path)},
            "stdout": exec_results.get("stdout", ""),
            "stderr": exec_results.get("stderr", ""),
            "language": detected_lang,
            "context_used": len(full_context)
        }
        
        return {
            "experimentCode": code_text, 
            "experimentResults": results, 
            "execution_path": path,
            **add_status_update("Experimenter", f"{lang_name} script created")
        }
        

def run_synthesis_agent(state: AgentState):
    log.info("--- SYNTHESIS ---")
    _state = state or {}
    path = ensure_list(_state, 'execution_path') + ["Synthesis"]

    exp_results = _state.get('experimentResults')
    pm_plan = _state.get('pmPlan', {}) or {}

    synthesis_context = [
        f"=== USER REQUEST ===\n{_state.get('userInput', '')}",
        f"\n=== OBJECTIVE ===\n{_state.get('coreObjectivePrompt', '')}",
        f"\n=== PLAN ===\n{json.dumps(pm_plan.get('plan_steps', []), indent=2)}",
    ]

    artifact_details = []
    artifact_message = ""

    if exp_results and isinstance(exp_results, dict):
        paths = exp_results.get("paths") or {}
    
        if paths:
            artifact_lines = []
            for artifact_type, artifact_path in paths.items():
                artifact_lines.append(f"- **{artifact_type.title()}**: `{os.path.basename(artifact_path)}`")
                artifact_details.append(f"{artifact_type}: {artifact_path}")
        
            artifact_message = "\n\n**Artifacts Generated:**\n" + "\n".join(artifact_lines)
            synthesis_context.append(f"\n=== ARTIFACTS ===\n" + "\n".join(artifact_details))
    
        if exp_results.get('stdout'):
            synthesis_context.append(f"\n=== OUTPUT ===\n{exp_results.get('stdout', '')}")
    
        if exp_results.get('stderr'):
            synthesis_context.append(f"\n=== ERRORS ===\n{exp_results.get('stderr', '')}")

    full_context = "\n".join(synthesis_context)

    synthesis_prompt = f"""Create FINAL RESPONSE after executing user's request.

{full_context}

Create comprehensive response that:
- Directly addresses original request
- Explains what was accomplished and HOW
- References specific artifacts and explains PURPOSE
- Provides context on how to USE deliverables
- Highlights KEY INSIGHTS
- Suggests NEXT STEPS if relevant
- Be SPECIFIC about what was created."""

    response = llm.invoke(synthesis_prompt)
    final_text = getattr(response, "content", "") or ""

    if artifact_message:
        final_text = final_text + "\n\n---\n" + artifact_message

    return {
        "draftResponse": final_text, 
        "execution_path": path,
        **add_status_update("Synthesis", "Response synthesized")
    }

def run_qa_agent(state: AgentState):
    log.info("--- QA ---")
    path = ensure_list(state, 'execution_path') + ["QA"]

    qa_context = [
        f"=== REQUEST ===\n{state.get('userInput', '')}",
        f"\n=== OBJECTIVE ===\n{state.get('coreObjectivePrompt', '')}",
        f"\n=== DRAFT ===\n{state.get('draftResponse', '')}",
    ]

    if state.get('experimentResults'):
        qa_context.append(f"\n=== ARTIFACTS ===\n{json.dumps(state.get('experimentResults', {}).get('paths', {}), indent=2)}")

    prompt = f"""You are a QA reviewer. Review the draft response against the user's objective.

{chr(10).join(qa_context)}

Review Instructions:
- Does the draft and its artifacts COMPLETELY satisfy ALL parts of the user's request?
- Is the quality of the work high?
- If this is a re-submission (rework cycle > 1), has the previous feedback been successfully addressed?

Response Format (required JSON or a single word 'APPROVED'):

Either return EXACTLY the single word:
  APPROVED

Or return JSON like:
{{
  "approved": false,
  "feedback": "Specific, actionable items to fix (bullet list or numbered).",
  "required_changes": ["..."]
}}
"""

    try:
        response = llm.invoke(prompt)
        content = getattr(response, "content", "") or ""
    except Exception as e:
        log.exception("QA LLM call failed: %s", e)
        return {
            "approved": False, 
            "qaFeedback": "QA LLM failed; manual review required.", 
            "execution_path": path,
            **add_status_update("QA", "QA failed")
        }

    # If LLM returned APPROVED word, treat as approved
    if "APPROVED" in content.strip().upper() and len(content.strip()) <= 20:
        return {
            "approved": True, 
            "qaFeedback": None, 
            "execution_path": path,
            **add_status_update("QA", "Approved")
        }

    # Else try JSON parse
    parsed = parse_json_from_llm(content)
    if isinstance(parsed, dict):
        approved = bool(parsed.get("approved", False))
        feedback = parsed.get("feedback") or parsed.get("qaFeedback") or parsed.get("required_changes") or ""
        # Normalize feedback to string
        if isinstance(feedback, list):
            feedback = "\n".join([str(x) for x in feedback])
        elif not isinstance(feedback, str):
            feedback = str(feedback)
        return {
            "approved": approved, 
            "qaFeedback": feedback if not approved else None, 
            "execution_path": path,
            **add_status_update("QA", "QA completed")
        }
    
    # Fallback: return raw text as feedback (not approved)
    safe_feedback = content.strip()[:2000] or "QA produced no actionable output."
    return {
        "approved": False, 
        "qaFeedback": safe_feedback, 
        "execution_path": path,
        **add_status_update("QA", "QA needs rework")
    }

def run_archivist_agent(state: AgentState):
    log.info("--- ARCHIVIST ---")
    path = ensure_list(state, 'execution_path') + ["Archivist"]

    summary_prompt = f"Summarize for memory.\n\nObjective: {state.get('coreObjectivePrompt')}\n\nResponse: {state.get('draftResponse')}\n\nSummary:"
    response = llm.invoke(summary_prompt)
    memory_manager.add_to_memory(getattr(response,"content",""), {"objective": state.get('coreObjectivePrompt')})

    return {
        "execution_path": path,
        **add_status_update("Archivist", "Saved to memory")
    }

def run_disclaimer_agent(state: AgentState):
    log.warning("--- DISCLAIMER ---")
    path = ensure_list(state, 'execution_path') + ["Disclaimer"]

    reason = "Budget limit reached." if state.get('budget_exceeded') else "Rework limit reached."
    disclaimer = f"**DISCLAIMER: {reason} Draft may be incomplete.**\n\n---\n\n"
    final_response = disclaimer + state.get('draftResponse', "No response")

    return {
        "draftResponse": final_response, 
        "execution_path": path,
        **add_status_update("Disclaimer", reason)
    }

def should_continue(state: AgentState):
    # Budget check first
    if state.get("budget_exceeded"):
        return "disclaimer_agent"
    try:
        rework = int(state.get("rework_cycles", 0))
        max_loops_allowed = int(state.get("max_loops", 0))
    except Exception:
        rework = state.get("rework_cycles", 0) or 0
        max_loops_allowed = state.get("max_loops", 0) or 0

    # If approved -> archive
    if state.get("approved"):
        return "archivist_agent"
    # If we have exceeded allowed reworks -> disclaimer
    if rework > max_loops_allowed:
        return "disclaimer_agent"
    # Default: return pm_agent so planner will create next plan
    return "pm_agent"

def should_run_experiment(state: AgentState):
    pm = state.get('pmPlan', {}) or {}
    return "experimenter_agent" if pm.get('experiment_needed') else "synthesis_agent"

#--- Build graphs ---
triage_workflow = StateGraph(AgentState)
triage_workflow.add_node("triage", run_triage_agent)
triage_workflow.set_entry_point("triage")
triage_workflow.add_edge("triage", END)
triage_app = triage_workflow.compile()

planner_workflow = StateGraph(AgentState)
planner_workflow.add_node("planner", run_planner_agent)
planner_workflow.set_entry_point("planner")
planner_workflow.add_edge("planner", END)
planner_app = planner_workflow.compile()

main_workflow = StateGraph(AgentState)
main_workflow.add_node("memory_retriever", run_memory_retrieval)
main_workflow.add_node("intent_agent", run_intent_agent)
main_workflow.add_node("pm_agent", run_pm_agent)
main_workflow.add_node("experimenter_agent", run_experimenter_agent)
main_workflow.add_node("synthesis_agent", run_synthesis_agent)
main_workflow.add_node("qa_agent", run_qa_agent)
main_workflow.add_node("archivist_agent", run_archivist_agent)
main_workflow.add_node("disclaimer_agent", run_disclaimer_agent)

main_workflow.set_entry_point("memory_retriever")
main_workflow.add_edge("memory_retriever", "intent_agent")
main_workflow.add_edge("intent_agent", "pm_agent")
main_workflow.add_edge("experimenter_agent", "synthesis_agent")
main_workflow.add_edge("synthesis_agent", "qa_agent")
main_workflow.add_edge("archivist_agent", END)
main_workflow.add_edge("disclaimer_agent", END)

main_workflow.add_conditional_edges("pm_agent", should_run_experiment)
main_workflow.add_conditional_edges("qa_agent", should_continue, {
    "archivist_agent": "archivist_agent",
    "pm_agent": "pm_agent",
    "disclaimer_agent": "disclaimer_agent"
})

main_app = main_workflow.compile()