diff --git a/agregar_planos.py b/agregar_planos.py
new file mode 100644
index 0000000..45f74b2
--- /dev/null
+++ b/agregar_planos.py
@@ -0,0 +1,621 @@
+from __future__ import annotations
+
+import os
+import sys
+import json
+import re
+import unicodedata
+import datetime as dt
+from decimal import Decimal
+from urllib.request import Request, urlopen
+
+# --------------------
+# Boot libs (Rocketbot portable)
+# --------------------
+base_dir = os.path.dirname(sys.executable)  # ...\Rocketbot
+libs_dir = os.path.join(base_dir, "py_libs", "py310")
+sys.path.insert(0, libs_dir)
+
+# Rocketbot a veces ya trae pyparsing viejo en memoria
+for k in list(sys.modules.keys()):
+    if k == "pyparsing" or k.startswith("pyparsing."):
+        del sys.modules[k]
+
+from googleapiclient.discovery import build
+from googleapiclient.errors import HttpError
+from google.oauth2 import service_account
+from google.oauth2.credentials import Credentials
+from google_auth_oauthlib.flow import InstalledAppFlow
+from google.auth.transport.requests import Request as GRequest
+
+# --------------------
+# Rocketbot vars helpers
+# --------------------
+def _missing(v) -> bool:
+    if v is None:
+        return True
+    if isinstance(v, str):
+        s = v.strip()
+        return s == "" or s == "ERROR_NOT_VAR"
+    return False
+
+def _gv(name, default=None):
+    try:
+        v = GetVar(name)
+    except Exception:
+        return default
+    return default if _missing(v) else v
+
+def _gvs(name, default="") -> str:
+    v = _gv(name, default)
+    if v is None:
+        return default
+    s = str(v).strip()
+    return default if (s == "" or s == "ERROR_NOT_VAR") else s
+
+def _sv(name, value):
+    try:
+        SetVar(name, value)
+    except Exception:
+        pass
+
+# --------------------
+# Google Auth
+# --------------------
+SCOPES = {scopes_api_google}
+
+def _load_json(path):
+    with open(path, "r", encoding="utf-8") as f:
+        return json.load(f)
+
+def get_services(credentials_json_path, token_json_path):
+    info = _load_json(credentials_json_path)
+
+    # Service Account
+    if isinstance(info, dict) and info.get("type") == "service_account":
+        creds = service_account.Credentials.from_service_account_file(
+            credentials_json_path, scopes=SCOPES
+        )
+        docs = build("docs", "v1", credentials=creds, cache_discovery=False)
+        drive = build("drive", "v3", credentials=creds, cache_discovery=False)
+        return docs, drive, "service_account"
+
+    # OAuth Desktop
+    creds = None
+    if os.path.exists(token_json_path):
+        creds = Credentials.from_authorized_user_file(token_json_path, SCOPES)
+
+    if (not creds) or (not creds.valid):
+        if creds and creds.expired and creds.refresh_token:
+            creds.refresh(GRequest())
+        else:
+            flow = InstalledAppFlow.from_client_secrets_file(credentials_json_path, SCOPES)
+            try:
+                creds = flow.run_local_server(port=0)
+            except Exception:
+                creds = flow.run_console()
+
+        os.makedirs(os.path.dirname(token_json_path), exist_ok=True)
+        with open(token_json_path, "w", encoding="utf-8") as f:
+            f.write(creds.to_json())
+
+    docs = build("docs", "v1", credentials=creds, cache_discovery=False)
+    drive = build("drive", "v3", credentials=creds, cache_discovery=False)
+    return docs, drive, "oauth"
+
+# --------------------
+# Drive helpers
+# --------------------
+def extract_doc_id_from_url(url):
+    m = re.search(r"/document/d/([a-zA-Z0-9_-]+)", url or "")
+    return m.group(1) if m else ""
+
+def ensure_docs_api_compatible(drive_service, file_id: str):
+    """
+    Si el ID es shortcut o NO es Google Docs nativo (ej: docx),
+    crea una copia convertida a application/vnd.google-apps.document y devuelve el nuevo ID.
+    """
+    meta = drive_service.files().get(
+        fileId=file_id,
+        fields="id,name,mimeType,shortcutDetails",
+        supportsAllDrives=True
+    ).execute()
+
+    # Shortcut -> target
+    if meta.get("mimeType") == "application/vnd.google-apps.shortcut":
+        target = (meta.get("shortcutDetails") or {}).get("targetId", "")
+        if not target:
+            raise RuntimeError("Es shortcut pero no trae targetId.")
+        file_id = target
+        meta = drive_service.files().get(
+            fileId=file_id,
+            fields="id,name,mimeType",
+            supportsAllDrives=True
+        ).execute()
+
+    # Convert if not native Google Doc
+    if meta.get("mimeType") != "application/vnd.google-apps.document":
+        new_name = (meta.get("name") or "Documento") + " (Google Docs)"
+        converted = drive_service.files().copy(
+            fileId=file_id,
+            body={"name": new_name, "mimeType": "application/vnd.google-apps.document"},
+            fields="id,name,mimeType",
+            supportsAllDrives=True
+        ).execute()
+        return converted["id"], converted, True
+
+    return file_id, meta, False
+
+# --------------------
+# Docs GET (tabs)
+# --------------------
+def docs_get(docs_service, doc_id: str):
+    # includeTabsContent=true para docs con Tabs
+    try:
+        return docs_service.documents().get(documentId=doc_id, includeTabsContent=True).execute()
+    except TypeError:
+        return docs_service.documents().get(documentId=doc_id).execute()
+
+def _walk_tabs(tabs):
+    if not tabs:
+        return
+    for t in tabs:
+        yield t
+        for x in _walk_tabs(t.get("childTabs") or []):
+            yield x
+
+def _get_tab_body_content(doc, tab_id):
+    tabs = doc.get("tabs")
+    if not tabs:
+        return (doc.get("body") or {}).get("content", [])
+    for t in _walk_tabs(tabs):
+        tid = ((t.get("tabProperties") or {}).get("tabId")) or ""
+        if tid == (tab_id or ""):
+            dtab = t.get("documentTab") or {}
+            return (dtab.get("body") or {}).get("content", [])
+    return []
+
+# --------------------
+# Marker search (BODY only, robust)
+# --------------------
+def _is_alnum(ch: str) -> bool:
+    return ch.isalnum()
+
+def _norm_alnum(s: str) -> str:
+    if s is None:
+        return ""
+    s = unicodedata.normalize("NFKD", str(s)).lower()
+    out = []
+    for ch in s:
+        if unicodedata.category(ch) == "Mn":
+            continue
+        if ch.isalnum():
+            out.append(ch)
+    return "".join(out)
+
+def _iter_text_chars_from_body(doc, tab_id):
+    """
+    Yields (docIndex:int, ch:str) en orden, solo del BODY.
+    """
+    content = _get_tab_body_content(doc, tab_id)
+
+    def walk(content_list):
+        for el in content_list or []:
+            if not isinstance(el, dict):
+                continue
+
+            p = el.get("paragraph")
+            if p:
+                for pe in p.get("elements", []):
+                    tr = pe.get("textRun")
+                    if not tr:
+                        continue
+                    txt = tr.get("content", "")
+                    st = pe.get("startIndex")
+                    if st is None or txt is None:
+                        continue
+                    for i, ch in enumerate(txt):
+                        yield (st + i, ch)
+                continue
+
+            t = el.get("table")
+            if t:
+                for row in t.get("tableRows", []):
+                    for cell in row.get("tableCells", []):
+                        yield from walk(cell.get("content", []))
+                continue
+
+            toc = el.get("tableOfContents")
+            if toc:
+                yield from walk(toc.get("content", []))
+                continue
+
+    yield from walk(content)
+
+def find_marker_in_body(doc, marker: str):
+    """
+    Busca marker en el BODY (no headers/footers), soporta tabs.
+    Matching: alfanumérico normalizado (fuzzy).
+    Retorna dict: {tabId, start, end, mode}
+    """
+    target = _norm_alnum(marker)
+    if not target:
+        return None
+
+    tabs = doc.get("tabs")
+    tab_ids = [None]
+    if tabs:
+        tab_ids = []
+        for t in _walk_tabs(tabs):
+            tid = ((t.get("tabProperties") or {}).get("tabId")) or ""
+            tab_ids.append(tid)
+
+    for tab_id in tab_ids:
+        norm_chars = []
+        norm_to_docidx = []  # norm position -> real doc index of that char
+
+        for doc_i, ch in _iter_text_chars_from_body(doc, tab_id):
+            if _is_alnum(ch):
+                norm_chars.append(ch.lower())
+                norm_to_docidx.append(doc_i)
+
+        norm_text = "".join(norm_chars)
+        pos = norm_text.find(target)
+        if pos != -1:
+            start_doc = norm_to_docidx[pos]
+            end_doc = norm_to_docidx[pos + len(target) - 1] + 1
+            return {"tabId": tab_id, "start": int(start_doc), "end": int(end_doc), "mode": "fuzzy_alnum"}
+
+    return None
+
+# --------------------
+# Docs batch helpers (tab-aware)
+# --------------------
+def _loc(index, tab_id):
+    d = {"index": int(index)}
+    if tab_id:
+        d["tabId"] = tab_id
+    return d
+
+def _range(start, end, tab_id):
+    d = {"startIndex": int(start), "endIndex": int(end)}
+    if tab_id:
+        d["tabId"] = tab_id
+    return d
+
+def _dim_pt(x: float):
+    return {"magnitude": float(x), "unit": "PT"}
+
+def _batch_update(docs_service, doc_id, reqs, chunk=900):
+    for i in range(0, len(reqs), chunk):
+        docs_service.documents().batchUpdate(
+            documentId=doc_id,
+            body={"requests": reqs[i:i + chunk]}
+        ).execute()
+
+# --------------------
+# Page box (max size)
+# --------------------
+def _get_pt(dimension_obj, default_val):
+    try:
+        if not dimension_obj:
+            return float(default_val)
+        mag = dimension_obj.get("magnitude", default_val)
+        return float(mag)
+    except Exception:
+        return float(default_val)
+
+def get_page_limits_pt(doc):
+    """
+    Devuelve (max_w_pt, max_h_pt) basado en pageSize - margins.
+    Fallback a Letter 8.5x11 con 1in margen si no existe.
+    """
+    ds = doc.get("documentStyle") or {}
+    ps = ds.get("pageSize") or {}
+    w_pt = _get_pt(ps.get("width"), 612.0)   # 8.5in * 72
+    h_pt = _get_pt(ps.get("height"), 792.0)  # 11in * 72
+
+    ml = _get_pt(ds.get("marginLeft"), 72.0)
+    mr = _get_pt(ds.get("marginRight"), 72.0)
+    mt = _get_pt(ds.get("marginTop"), 72.0)
+    mb = _get_pt(ds.get("marginBottom"), 72.0)
+
+    max_w = max(100.0, w_pt - ml - mr - 12.0)
+    max_h = max(100.0, h_pt - mt - mb - 12.0)
+    return max_w, max_h
+
+# --------------------
+# Image URL + dimensions
+# --------------------
+def extract_drive_file_id(url: str) -> str:
+    if not url:
+        return ""
+    # /file/d/<id>/... or /d/<id>/...
+    m = re.search(r"/d/([a-zA-Z0-9_-]{10,})", url)
+    if m:
+        return m.group(1)
+    # ?id=<id>
+    m = re.search(r"[?&]id=([a-zA-Z0-9_-]{10,})", url)
+    if m:
+        return m.group(1)
+    return ""
+
+def sniff_png_wh(data: bytes):
+    # PNG: width/height en IHDR (bytes 16..24)
+    if len(data) >= 24 and data[:8] == b"\x89PNG\r\n\x1a\n":
+        w = int.from_bytes(data[16:20], "big")
+        h = int.from_bytes(data[20:24], "big")
+        return w, h
+    return None
+
+def sniff_gif_wh(data: bytes):
+    if len(data) >= 10 and (data[:6] in (b"GIF87a", b"GIF89a")):
+        w = int.from_bytes(data[6:8], "little")
+        h = int.from_bytes(data[8:10], "little")
+        return w, h
+    return None
+
+def sniff_jpeg_wh(data: bytes):
+    # JPEG: buscar SOF0/SOF2
+    if len(data) < 4 or data[0:2] != b"\xFF\xD8":
+        return None
+    i = 2
+    n = len(data)
+    while i + 9 < n:
+        if data[i] != 0xFF:
+            i += 1
+            continue
+        # saltar FFs
+        while i < n and data[i] == 0xFF:
+            i += 1
+        if i >= n:
+            break
+        marker = data[i]
+        i += 1
+        # markers sin length
+        if marker in (0xD8, 0xD9):
+            continue
+        if i + 1 >= n:
+            break
+        seglen = int.from_bytes(data[i:i+2], "big")
+        if seglen < 2 or i + seglen > n:
+            break
+        # SOF0..SOF3, SOF5..SOF7, SOF9..SOF11, SOF13..SOF15
+        if marker in (0xC0, 0xC1, 0xC2, 0xC3, 0xC5, 0xC6, 0xC7, 0xC9, 0xCA, 0xCB, 0xCD, 0xCE, 0xCF):
+            # data: [len2][precision1][height2][width2]...
+            if i + 7 < n:
+                h = int.from_bytes(data[i+3:i+5], "big")
+                w = int.from_bytes(data[i+5:i+7], "big")
+                return w, h
+        i += seglen
+    return None
+
+def sniff_image_wh_from_url(url: str, timeout=20):
+    """
+    Intenta leer solo cabecera (Range) para detectar dims.
+    """
+    headers = {
+        "User-Agent": "Mozilla/5.0",
+        "Range": "bytes=0-65535",
+    }
+    req = Request(url, headers=headers)
+    with urlopen(req, timeout=timeout) as resp:
+        data = resp.read(65536)
+
+    for fn in (sniff_png_wh, sniff_gif_wh, sniff_jpeg_wh):
+        wh = fn(data)
+        if wh:
+            return wh[0], wh[1]
+    return None, None
+
+def resolve_image(drive_service, url: str):
+    """
+    Retorna: (uri_para_docs, w_px, h_px, source)
+    Si es link de Drive y NO se puede acceder/leer metadata, retorna uri="" (no encontrado).
+    """
+    if not url:
+        return ("", None, None, "empty")
+
+    fid = extract_drive_file_id(url)
+    if fid:
+        try:
+            meta = drive_service.files().get(
+                fileId=fid,
+                fields="id,mimeType,imageMediaMetadata(width,height)",
+                supportsAllDrives=True
+            ).execute()
+
+            im = meta.get("imageMediaMetadata") or {}
+            w = im.get("width")
+            h = im.get("height")
+
+            uri = f"https://drive.google.com/uc?export=download&id={fid}"
+            return (uri, w, h, "drive")
+
+        except Exception as e:
+            return ("", None, None, f"drive_error:{type(e).__name__}")
+
+    # URL externa
+    w = h = None
+    try:
+        w, h = sniff_image_wh_from_url(url)
+    except Exception:
+        w, h = None, None
+
+    return (url, w, h, "url")
+
+def fit_object_size_pt(w_px, h_px, max_w_pt, max_h_pt):
+    """
+    Convierte px->pt usando 0.75 (96dpi->72pt) y escala para no exceder max_w/max_h.
+    """
+    if not w_px or not h_px:
+        # fallback seguro
+        w_pt = float(max_w_pt)
+        h_pt = float(min(max_h_pt, max_w_pt * 0.75))
+        return w_pt, h_pt
+
+    w_pt0 = float(w_px) * 0.75
+    h_pt0 = float(h_px) * 0.75
+    if w_pt0 <= 0 or h_pt0 <= 0:
+        w_pt = float(max_w_pt)
+        h_pt = float(min(max_h_pt, max_w_pt * 0.75))
+        return w_pt, h_pt
+
+    scale = min(max_w_pt / w_pt0, max_h_pt / h_pt0, 1.0)
+    return w_pt0 * scale, h_pt0 * scale
+
+# --------------------
+# Insert planos at marker
+# --------------------
+def insert_planos_at_marker(docs_service, drive_service, doc_id: str, marker: str, url1: str, url2: str):
+    # Flags Rocketbot: arrancar siempre en False
+    _sv("plano1_found", "False")
+    _sv("plano2_found", "False")
+
+    doc = docs_get(docs_service, doc_id)
+
+    hit = find_marker_in_body(doc, marker)
+    if not hit:
+        _sv("gdoc_planos_marker_found", "0")
+        _sv("gdoc_planos_marker", marker)
+        raise RuntimeError("No encontre el marcador en el BODY del documento: " + str(marker))
+
+    tab_id = hit.get("tabId")
+    start = int(hit["start"])
+    end = int(hit["end"])
+
+    _sv("gdoc_planos_marker_found", "1")
+    _sv("gdoc_planos_marker_mode", hit.get("mode", ""))
+    _sv("gdoc_planos_marker_tabId", tab_id or "")
+    _sv("gdoc_planos_marker_start", str(start))
+    _sv("gdoc_planos_marker_end", str(end))
+
+    max_w_pt, max_h_pt = get_page_limits_pt(doc)
+    _sv("gdoc_planos_max_w_pt", str(max_w_pt))
+    _sv("gdoc_planos_max_h_pt", str(max_h_pt))
+
+    # Resolver URLs + dims (y setear flags)
+    uri1, w1, h1, src1 = resolve_image(drive_service, url1)  # plano producto
+    if uri1:
+        _sv("plano1_found", "True")
+
+    uri2, w2, h2, src2 = resolve_image(drive_service, url2)  # plano piso
+    if uri2:
+        _sv("plano2_found", "True")
+
+    # Validación: si falta alguno, cortar (y Rocketbot ya ve cuáles quedaron True/False)
+    if not uri1 or not uri2:
+        if not uri1 and not uri2:
+            raise RuntimeError("No se encontró ningún plano (plano 1 y plano 2). Revisá url_plano_producto y url_plano_piso.")
+        if not uri1:
+            raise RuntimeError("No se encontró el plano 1 (url_plano_producto). Verificá la URL / permisos en Drive.")
+        raise RuntimeError("No se encontró el plano 2 (url_plano_piso). Verificá la URL / permisos en Drive.")
+
+    _sv("gdoc_planos_uri1", uri1)
+    _sv("gdoc_planos_uri2", uri2)
+    _sv("gdoc_planos_src1", src1)
+    _sv("gdoc_planos_src2", src2)
+    _sv("gdoc_planos_w1_px", "" if w1 is None else str(w1))
+    _sv("gdoc_planos_h1_px", "" if h1 is None else str(h1))
+    _sv("gdoc_planos_w2_px", "" if w2 is None else str(w2))
+    _sv("gdoc_planos_h2_px", "" if h2 is None else str(h2))
+
+    w1_pt, h1_pt = fit_object_size_pt(w1, h1, max_w_pt, max_h_pt)
+    w2_pt, h2_pt = fit_object_size_pt(w2, h2, max_w_pt, max_h_pt)
+
+    # Construcción robusta SIN depender de index+1:
+    # Insertamos SIEMPRE en el mismo start para apilar:
+    # [img1][\n][img2][\n]
+    reqs = [
+        {"deleteContentRange": {"range": _range(start, end, tab_id)}},
+
+        # 1) newline final
+        {"insertText": {"location": _loc(start, tab_id), "text": "\n"}},
+
+        # 2) img2 (queda antes del newline final)
+        {"insertInlineImage": {
+            "location": _loc(start, tab_id),
+            "uri": uri2,
+            "objectSize": {"width": _dim_pt(w2_pt), "height": _dim_pt(h2_pt)},
+        }},
+
+        # 3) newline entre imágenes (queda antes de img2)
+        {"insertText": {"location": _loc(start, tab_id), "text": "\n"}},
+
+        # 4) img1 (queda antes de newline entre imágenes)
+        {"insertInlineImage": {
+            "location": _loc(start, tab_id),
+            "uri": uri1,
+            "objectSize": {"width": _dim_pt(w1_pt), "height": _dim_pt(h1_pt)},
+        }},
+    ]
+
+    _batch_update(docs_service, doc_id, reqs)
+
+# Rocketbot quirk FIX
+globals().update(locals())
+
+# --------------------
+# MAIN
+# --------------------
+try:
+    # Flags globales: arrancar siempre en False (por si fallamos antes de la función)
+    _sv("plano1_found", "False")
+    _sv("plano2_found", "False")
+
+    marker = _gvs("gdoc_marker_planos", "INSERTAR_PLANOS")
+
+    url1 = _gvs("url_plano_producto", "")
+    url2 = _gvs("url_plano_piso", "")
+
+    cred_path = _gvs("gdoc_sa_json", "")
+    if cred_path == "":
+        raise RuntimeError("Falta gdoc_sa_json (ruta a credentials.json)")
+    if not os.path.isabs(cred_path):
+        cred_path = os.path.join(base_dir, cred_path)
+    if not os.path.exists(cred_path):
+        raise RuntimeError("No existe credentials.json: " + cred_path)
+
+    token_path = _gvs("gdoc_token_json", "")
+    if token_path == "":
+        token_path = os.path.join(base_dir, "credentials", "token_gdocs_drive.json")
+    if not os.path.isabs(token_path):
+        token_path = os.path.join(base_dir, token_path)
+
+    _sv("gdoc_credentials_used", cred_path)
+    _sv("gdoc_token_used", token_path)
+
+    docs_service, drive_service, auth_mode = get_services(cred_path, token_path)
+    _sv("gdoc_auth_mode", auth_mode)
+
+    # doc_id desde URL del navegador
+    url_doc = "{current_url}"
+    #if url_doc == "":
+    #    url_doc = _gvs("current_url", "")
+    #_sv("gdoc_url_used", url_doc)
+
+    doc_id = extract_doc_id_from_url(url_doc)
+    if doc_id == "":
+        raise RuntimeError("No pude extraer documentId de la URL: " + url_doc)
+
+    doc_id_final, meta_file, converted = ensure_docs_api_compatible(drive_service, doc_id)
+    _sv("gdoc_original_id", doc_id)
+    _sv("gdoc_id", doc_id_final)
+    _sv("gdoc_file_name", meta_file.get("name", ""))
+    _sv("gdoc_file_mimeType", meta_file.get("mimeType", ""))
+    _sv("gdoc_converted", "1" if converted else "0")
+    _sv("gdoc_url_final", f"https://docs.google.com/document/d/{doc_id_final}/edit")
+
+    insert_planos_at_marker(docs_service, drive_service, doc_id_final, marker, url1, url2)
+
+    _sv("gdoc_planos_status", "OK")
+    _sv("gdoc_planos_error", "")
+
+except HttpError as e:
+    _sv("gdoc_planos_status", "ERROR")
+    _sv("gdoc_planos_error", "HttpError: " + str(e))
+    raise
+except Exception as e:
+    _sv("gdoc_planos_status", "ERROR")
+    _sv("gdoc_planos_error", str(e))
+    raise
\ No newline at end of file