jkj

cmdlet/add_file.py

@@ -6,6 +6,7 @@ import sys
 import shutil
 import tempfile
 import re
+from urllib.parse import urlsplit, parse_qs
 
 import models
 import pipeline as ctx
@@ -13,12 +14,20 @@ from API import HydrusNetwork as hydrus_wrapper
 from SYS.logger import log, debug
 from SYS.utils_constant import ALL_SUPPORTED_EXTENSIONS
 from Store import Store
-from ._shared import (
-    Cmdlet, CmdletArg, parse_cmdlet_args, SharedArgs,
-    extract_tag_from_result, extract_title_from_result, extract_url_from_result,
-    merge_sequences, extract_relationships, extract_duration, coerce_to_pipe_object
-)
-from ._shared import collapse_namespace_tag
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+parse_cmdlet_args = sh.parse_cmdlet_args
+SharedArgs = sh.SharedArgs
+extract_tag_from_result = sh.extract_tag_from_result
+extract_title_from_result = sh.extract_title_from_result
+extract_url_from_result = sh.extract_url_from_result
+merge_sequences = sh.merge_sequences
+extract_relationships = sh.extract_relationships
+extract_duration = sh.extract_duration
+coerce_to_pipe_object = sh.coerce_to_pipe_object
+collapse_namespace_tag = sh.collapse_namespace_tag
 from API.folder import read_sidecar, find_sidecar, write_sidecar, API_folder_store
 from SYS.utils import sha256_file, unique_path
 from metadata import write_metadata
@@ -181,7 +190,7 @@ class Add_File(Cmdlet):
                         downloaded_path = Path(downloaded)
                         if downloaded_path.exists() and downloaded_path.is_dir():
                             log(
-                                "[add-file] OpenLibrary download produced a directory (missing img2pdf?). Cannot ingest.",
+                                "[add-file] OpenLibrary download produced a directory (PDF conversion failed). Cannot ingest.",
                                 file=sys.stderr,
                             )
                             failures += 1
@@ -195,12 +204,32 @@ class Add_File(Cmdlet):
                     if isinstance(media_path_or_url, str) and media_path_or_url.lower().startswith(
                         ("http://", "https://", "magnet:", "torrent:")
                     ):
-                        code = self._delegate_to_download_data(item, media_path_or_url, location, provider_name, args, config)
-                        if code == 0:
-                            successes += 1
-                        else:
-                            failures += 1
-                        continue
+                        # Hydrus file URLs are direct file downloads and may require Hydrus auth headers.
+                        # If the user provided a destination (-provider or -store), download now and continue.
+                        if (provider_name or location) and isinstance(media_path_or_url, str) and media_path_or_url.lower().startswith(("http://", "https://")):
+                            downloaded = self._try_download_hydrus_file_url(
+                                file_url=str(media_path_or_url),
+                                pipe_obj=pipe_obj,
+                                config=config,
+                            )
+                            if downloaded is not None:
+                                downloaded_path, downloaded_temp_dir = downloaded
+                                temp_dir_to_cleanup = downloaded_temp_dir
+                                media_path_or_url = str(downloaded_path)
+                                pipe_obj.path = str(downloaded_path)
+                                pipe_obj.is_temp = True
+                                delete_after_item = True
+
+                        # If it's still a URL target, fall back to the legacy delegate.
+                        if isinstance(media_path_or_url, str) and media_path_or_url.lower().startswith(
+                            ("http://", "https://", "magnet:", "torrent:")
+                        ):
+                            code = self._delegate_to_download_data(item, media_path_or_url, location, provider_name, args, config)
+                            if code == 0:
+                                successes += 1
+                            else:
+                                failures += 1
+                            continue
 
                 media_path = Path(media_path_or_url) if isinstance(media_path_or_url, str) else media_path_or_url
 
@@ -767,6 +796,134 @@ class Add_File(Cmdlet):
             return True
         return False
 
+    @staticmethod
+    def _sanitize_filename(value: str) -> str:
+        # Minimal Windows-safe filename sanitization.
+        text = str(value or "").strip()
+        if not text:
+            return "file"
+        invalid = '<>:"/\\|?*'
+        text = "".join("_" if (ch in invalid or ord(ch) < 32) else ch for ch in text)
+        text = re.sub(r"\s+", " ", text).strip(" .")
+        return text or "file"
+
+    @staticmethod
+    def _parse_hydrus_file_url(file_url: str) -> Optional[str]:
+        """Return the sha256 hash from a Hydrus /get_files/file URL, or None."""
+        try:
+            split = urlsplit(str(file_url))
+            if split.scheme.lower() not in {"http", "https"}:
+                return None
+            path_lower = (split.path or "").lower()
+            if "/get_files/file" not in path_lower:
+                return None
+            params = parse_qs(split.query or "")
+            raw = None
+            if "hash" in params and params["hash"]:
+                raw = params["hash"][0]
+            if not raw:
+                return None
+            hash_val = str(raw).strip().lower()
+            if not re.fullmatch(r"[0-9a-f]{64}", hash_val):
+                return None
+            return hash_val
+        except Exception:
+            return None
+
+    def _try_download_hydrus_file_url(
+        self,
+        *,
+        file_url: str,
+        pipe_obj: models.PipeObject,
+        config: Dict[str, Any],
+    ) -> Optional[tuple[Path, Path]]:
+        """If *file_url* is a Hydrus file URL, download it to temp and return (path, temp_dir)."""
+        file_hash = self._parse_hydrus_file_url(file_url)
+        if not file_hash:
+            return None
+
+        # Resolve Hydrus backend for auth.
+        store_name = str(getattr(pipe_obj, "store", "") or "").strip()
+        if ":" in store_name:
+            store_name = store_name.split(":", 1)[-1].strip()
+
+        backend = None
+        try:
+            store_registry = Store(config)
+            if store_name and store_registry.is_available(store_name):
+                candidate = store_registry[store_name]
+                if type(candidate).__name__.lower() == "hydrusnetwork":
+                    backend = candidate
+        except Exception:
+            backend = None
+
+        if backend is None:
+            try:
+                store_registry = Store(config)
+                target_prefix = str(file_url).split("/get_files/file", 1)[0].rstrip("/")
+                for backend_name in store_registry.list_backends():
+                    candidate = store_registry[backend_name]
+                    if type(candidate).__name__.lower() != "hydrusnetwork":
+                        continue
+                    base_url = str(getattr(candidate, "URL", "") or "").rstrip("/")
+                    if base_url and (target_prefix.lower() == base_url.lower() or target_prefix.lower().startswith(base_url.lower())):
+                        backend = candidate
+                        break
+            except Exception:
+                backend = None
+
+        if backend is None:
+            debug("[add-file] Hydrus file URL detected but no Hydrus backend matched for auth")
+            return None
+
+        api_key = str(getattr(backend, "API", "") or "").strip()
+        if not api_key:
+            debug(f"[add-file] Hydrus backend '{getattr(backend, 'NAME', '') or store_name}' missing API key")
+            return None
+
+        # Best-effort filename from title + ext.
+        ext = ""
+        try:
+            if isinstance(pipe_obj.extra, dict):
+                ext = str(pipe_obj.extra.get("ext") or "").strip().lstrip(".")
+        except Exception:
+            ext = ""
+        if not ext:
+            ext = "bin"
+
+        title_hint = str(getattr(pipe_obj, "title", "") or "").strip()
+        base_name = self._sanitize_filename(title_hint) if title_hint else f"hydrus_{file_hash[:12]}"
+
+        temp_dir = Path(tempfile.mkdtemp(prefix="medios_hydrus_"))
+        destination = unique_path(temp_dir / f"{base_name}.{ext}")
+
+        headers = {"Hydrus-Client-API-Access-Key": api_key}
+        timeout = 60.0
+        try:
+            client = getattr(backend, "_client", None)
+            timeout_val = getattr(client, "timeout", None)
+            if timeout_val is not None:
+                timeout = float(timeout_val)
+        except Exception:
+            timeout = 60.0
+
+        try:
+            log(
+                f"[add-file] Downloading Hydrus file via API ({getattr(backend, 'NAME', '') or store_name})",
+                file=sys.stderr,
+            )
+            downloaded_bytes = hydrus_wrapper.download_hydrus_file(str(file_url), headers, destination, timeout)
+            if downloaded_bytes <= 0 and not destination.exists():
+                return None
+            return destination, temp_dir
+        except Exception as exc:
+            log(f"[add-file] Hydrus download failed: {exc}", file=sys.stderr)
+            try:
+                shutil.rmtree(temp_dir, ignore_errors=True)
+            except Exception:
+                pass
+            return None
+
     def _delegate_to_download_data(
         self,
         result: Any,
@@ -883,6 +1040,61 @@ class Add_File(Cmdlet):
         except Exception:
             return None
 
+    @staticmethod
+    def _get_note_text(result: Any, pipe_obj: models.PipeObject, note_name: str) -> Optional[str]:
+        """Extract a named note text from a piped item.
+
+        Supports:
+        - pipe_obj.extra["notes"][note_name]
+        - result["notes"][note_name] for dict results
+        - pipe_obj.extra[note_name] / result[note_name] as fallback
+        """
+
+        def _normalize(val: Any) -> Optional[str]:
+            if val is None:
+                return None
+            if isinstance(val, bytes):
+                try:
+                    val = val.decode("utf-8", errors="ignore")
+                except Exception:
+                    val = str(val)
+            if isinstance(val, str):
+                text = val.strip()
+                return text if text else None
+            try:
+                text = str(val).strip()
+                return text if text else None
+            except Exception:
+                return None
+
+        note_key = str(note_name or "").strip()
+        if not note_key:
+            return None
+
+        # Prefer notes dict on PipeObject.extra (common for cmdlet-emitted dicts)
+        try:
+            if isinstance(pipe_obj.extra, dict):
+                notes_val = pipe_obj.extra.get("notes")
+                if isinstance(notes_val, dict) and note_key in notes_val:
+                    return _normalize(notes_val.get(note_key))
+                if note_key in pipe_obj.extra:
+                    return _normalize(pipe_obj.extra.get(note_key))
+        except Exception:
+            pass
+
+        # Fallback to raw result dict
+        if isinstance(result, dict):
+            try:
+                notes_val = result.get("notes")
+                if isinstance(notes_val, dict) and note_key in notes_val:
+                    return _normalize(notes_val.get(note_key))
+                if note_key in result:
+                    return _normalize(result.get(note_key))
+            except Exception:
+                pass
+
+        return None
+
     @staticmethod
     def _update_pipe_object_destination(
         pipe_obj: models.PipeObject,
@@ -1451,6 +1663,26 @@ class Add_File(Cmdlet):
                 except Exception:
                     pass
 
+            # If a subtitle note was provided upstream (e.g., download-media writes notes.sub),
+            # persist it automatically like add-note would.
+            sub_note = Add_File._get_note_text(result, pipe_obj, "sub")
+            if sub_note:
+                try:
+                    setter = getattr(backend, "set_note", None)
+                    if callable(setter):
+                        setter(resolved_hash, "sub", sub_note)
+                except Exception:
+                    pass
+
+            chapters_note = Add_File._get_note_text(result, pipe_obj, "chapters")
+            if chapters_note:
+                try:
+                    setter = getattr(backend, "set_note", None)
+                    if callable(setter):
+                        setter(resolved_hash, "chapters", chapters_note)
+                except Exception:
+                    pass
+
             meta: Dict[str, Any] = {}
             try:
                 meta = backend.get_metadata(resolved_hash) or {}

cmdlet/add_note.py

@@ -7,15 +7,15 @@ import sys
 from SYS.logger import log
 
 import pipeline as ctx
-from ._shared import (
-    Cmdlet,
-    CmdletArg,
-    SharedArgs,
-    normalize_hash,
-    parse_cmdlet_args,
-    normalize_result_input,
-    should_show_help,
-)
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+normalize_hash = sh.normalize_hash
+parse_cmdlet_args = sh.parse_cmdlet_args
+normalize_result_input = sh.normalize_result_input
+should_show_help = sh.should_show_help
 from Store import Store
 from SYS.utils import sha256_file
 
@@ -84,9 +84,9 @@ class Add_Note(Cmdlet):
         else:
             note_text = str(text_parts or "").strip()
 
-        if not note_text:
-            log("[add_note] Error: Empty note text", file=sys.stderr)
-            return 1
+        # Note text can be omitted when upstream stages provide it (e.g. download-media --write-sub
+        # attaches notes.sub). In that case we resolve per-item below.
+        user_provided_text = bool(note_text)
 
         results = normalize_result_input(result)
         if not results:
@@ -99,11 +99,56 @@ class Add_Note(Cmdlet):
         store_registry = Store(config)
         updated = 0
 
+        # Optional global fallback for note text from pipeline values.
+        # Allows patterns like: ... | add-note sub
+        pipeline_default_text = None
+        if not user_provided_text:
+            try:
+                pipeline_default_text = ctx.load_value(note_name)
+            except Exception:
+                pipeline_default_text = None
+            if isinstance(pipeline_default_text, list):
+                pipeline_default_text = " ".join([str(x) for x in pipeline_default_text]).strip()
+            elif pipeline_default_text is not None:
+                pipeline_default_text = str(pipeline_default_text).strip()
+
         for res in results:
             if not isinstance(res, dict):
                 ctx.emit(res)
                 continue
 
+            # Resolve note text for this item when not provided explicitly.
+            item_note_text = note_text
+            if not user_provided_text:
+                # Prefer item-scoped notes dict.
+                candidate = None
+                try:
+                    notes = res.get("notes")
+                    if isinstance(notes, dict):
+                        candidate = notes.get(note_name)
+                except Exception:
+                    candidate = None
+
+                # Also allow direct field fallback: res["sub"], etc.
+                if candidate is None:
+                    try:
+                        candidate = res.get(note_name)
+                    except Exception:
+                        candidate = None
+
+                if candidate is None:
+                    candidate = pipeline_default_text
+
+                if isinstance(candidate, list):
+                    item_note_text = " ".join([str(x) for x in candidate]).strip()
+                else:
+                    item_note_text = str(candidate or "").strip()
+
+            if not item_note_text:
+                log(f"[add_note] Warning: No note text found for '{note_name}'; skipping", file=sys.stderr)
+                ctx.emit(res)
+                continue
+
             store_name = str(store_override or res.get("store") or "").strip()
             raw_hash = res.get("hash")
             raw_path = res.get("path")
@@ -130,7 +175,7 @@ class Add_Note(Cmdlet):
 
             ok = False
             try:
-                ok = bool(backend.set_note(resolved_hash, note_name, note_text, config=config))
+                ok = bool(backend.set_note(resolved_hash, note_name, item_note_text, config=config))
             except Exception as exc:
                 log(f"[add_note] Error: Failed to set note: {exc}", file=sys.stderr)
                 ok = False

cmdlet/add_relationship.py

@@ -11,7 +11,15 @@ from SYS.logger import log
 
 import pipeline as ctx
 from API import HydrusNetwork as hydrus_wrapper
-from ._shared import Cmdlet, CmdletArg, SharedArgs, parse_cmdlet_args, normalize_result_input, should_show_help, get_field
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+parse_cmdlet_args = sh.parse_cmdlet_args
+normalize_result_input = sh.normalize_result_input
+should_show_help = sh.should_show_help
+get_field = sh.get_field
 from API.folder import read_sidecar, find_sidecar, API_folder_store
 from Store import Store
 

cmdlet/add_tag.py

@@ -8,19 +8,20 @@ from SYS.logger import log
 
 import models
 import pipeline as ctx
-from ._shared import normalize_result_input, filter_results_by_temp
-from ._shared import (
-    Cmdlet,
-    CmdletArg,
-    SharedArgs,
-    normalize_hash,
-    parse_tag_arguments,
-    expand_tag_groups,
-    parse_cmdlet_args,
-    collapse_namespace_tag,
-    should_show_help,
-    get_field,
-)
+from . import _shared as sh
+
+normalize_result_input = sh.normalize_result_input
+filter_results_by_temp = sh.filter_results_by_temp
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+normalize_hash = sh.normalize_hash
+parse_tag_arguments = sh.parse_tag_arguments
+expand_tag_groups = sh.expand_tag_groups
+parse_cmdlet_args = sh.parse_cmdlet_args
+collapse_namespace_tag = sh.collapse_namespace_tag
+should_show_help = sh.should_show_help
+get_field = sh.get_field
 from Store import Store
 from SYS.utils import sha256_file
 

cmdlet/add_tags.py

@@ -8,19 +8,20 @@ from SYS.logger import log
 
 import models
 import pipeline as ctx
-from ._shared import normalize_result_input, filter_results_by_temp
-from ._shared import (
-    Cmdlet,
-    CmdletArg,
-    SharedArgs,
-    normalize_hash,
-    parse_tag_arguments,
-    expand_tag_groups,
-    parse_cmdlet_args,
-    collapse_namespace_tags,
-    should_show_help,
-    get_field,
-)
+from . import _shared as sh
+
+normalize_result_input = sh.normalize_result_input
+filter_results_by_temp = sh.filter_results_by_temp
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+normalize_hash = sh.normalize_hash
+parse_tag_arguments = sh.parse_tag_arguments
+expand_tag_groups = sh.expand_tag_groups
+parse_cmdlet_args = sh.parse_cmdlet_args
+collapse_namespace_tags = sh.collapse_namespace_tags
+should_show_help = sh.should_show_help
+get_field = sh.get_field
 from Store import Store
 from SYS.utils import sha256_file
 

cmdlet/add_url.py

@@ -4,12 +4,12 @@ from typing import Any, Dict, Sequence
 import sys
 
 import pipeline as ctx
-from ._shared import Cmdlet, CmdletArg, SharedArgs, parse_cmdlet_args, get_field, normalize_hash
+from . import _shared as sh
 from SYS.logger import log
 from Store import Store
 
 
-class Add_Url(Cmdlet):
+class Add_Url(sh.Cmdlet):
     """Add URL associations to files via hash+store."""
 
     def __init__(self) -> None:
@@ -18,9 +18,9 @@ class Add_Url(Cmdlet):
             summary="Associate a URL with a file",
             usage="@1 | add-url <url>",
             arg=[
-                SharedArgs.HASH,
-                SharedArgs.STORE,
-                CmdletArg("url", required=True, description="URL to associate"),
+                sh.SharedArgs.HASH,
+                sh.SharedArgs.STORE,
+                sh.CmdletArg("url", required=True, description="URL to associate"),
             ],
             detail=[
                 "- Associates URL with file identified by hash+store",
@@ -32,11 +32,11 @@ class Add_Url(Cmdlet):
     
     def run(self, result: Any, args: Sequence[str], config: Dict[str, Any]) -> int:
         """Add URL to file via hash+store backend."""
-        parsed = parse_cmdlet_args(args, self)
+        parsed = sh.parse_cmdlet_args(args, self)
         
         # Extract hash and store from result or args
-        file_hash = parsed.get("hash") or get_field(result, "hash")
-        store_name = parsed.get("store") or get_field(result, "store")
+        file_hash = parsed.get("hash") or sh.get_field(result, "hash")
+        store_name = parsed.get("store") or sh.get_field(result, "store")
         url_arg = parsed.get("url")
         
         if not file_hash:
@@ -52,7 +52,7 @@ class Add_Url(Cmdlet):
             return 1
         
         # Normalize hash
-        file_hash = normalize_hash(file_hash)
+        file_hash = sh.normalize_hash(file_hash)
         if not file_hash:
             log("Error: Invalid hash format")
             return 1

cmdlet/catalog.py

@@ -1,215 +0,0 @@
-from __future__ import annotations
-
-from importlib import import_module
-from typing import Any, Dict, List, Optional
-
-try:
-    from cmdlet import REGISTRY
-except Exception:
-    REGISTRY = {}  # type: ignore
-
-try:
-    from cmdnat import register_native_commands as _register_native_commands
-except Exception:
-    _register_native_commands = None
-
-
-def ensure_registry_loaded() -> None:
-    """Ensure native commands are registered into REGISTRY (idempotent)."""
-    if _register_native_commands and REGISTRY is not None:
-        try:
-            _register_native_commands(REGISTRY)
-        except Exception:
-            pass
-
-
-def _normalize_mod_name(mod_name: str) -> str:
-    """Normalize a command/module name for import resolution."""
-    normalized = (mod_name or "").strip()
-    if normalized.startswith('.'):
-        normalized = normalized.lstrip('.')
-    normalized = normalized.replace('-', '_')
-    return normalized
-
-
-def import_cmd_module(mod_name: str):
-    """Import a cmdlet/native module from cmdnat or cmdlet packages."""
-    normalized = _normalize_mod_name(mod_name)
-    if not normalized:
-        return None
-    for package in ("cmdnat", "cmdlet", None):
-        try:
-            qualified = f"{package}.{normalized}" if package else normalized
-            return import_module(qualified)
-        except ModuleNotFoundError:
-            continue
-        except Exception:
-            continue
-    return None
-
-
-def _normalize_arg(arg: Any) -> Dict[str, Any]:
-    """Convert a CmdletArg/dict into a plain metadata dict."""
-    if isinstance(arg, dict):
-        name = arg.get("name", "")
-        return {
-            "name": str(name).lstrip("-"),
-            "type": arg.get("type", "string"),
-            "required": bool(arg.get("required", False)),
-            "description": arg.get("description", ""),
-            "choices": arg.get("choices", []) or [],
-            "alias": arg.get("alias", ""),
-            "variadic": arg.get("variadic", False),
-        }
-
-    name = getattr(arg, "name", "") or ""
-    return {
-        "name": str(name).lstrip("-"),
-        "type": getattr(arg, "type", "string"),
-        "required": bool(getattr(arg, "required", False)),
-        "description": getattr(arg, "description", ""),
-        "choices": getattr(arg, "choices", []) or [],
-        "alias": getattr(arg, "alias", ""),
-        "variadic": getattr(arg, "variadic", False),
-    }
-
-
-def get_cmdlet_metadata(cmd_name: str) -> Optional[Dict[str, Any]]:
-    """Return normalized metadata for a cmdlet, if available (aliases supported)."""
-    ensure_registry_loaded()
-    normalized = cmd_name.replace("-", "_")
-    mod = import_cmd_module(normalized)
-    data = getattr(mod, "CMDLET", None) if mod else None
-
-    if data is None:
-        try:
-            reg_fn = (REGISTRY or {}).get(cmd_name.replace('_', '-').lower())
-            if reg_fn:
-                owner_mod = getattr(reg_fn, "__module__", "")
-                if owner_mod:
-                    owner = import_module(owner_mod)
-                    data = getattr(owner, "CMDLET", None)
-        except Exception:
-            data = None
-
-    if not data:
-        return None
-
-    if hasattr(data, "to_dict"):
-        base = data.to_dict()
-    elif isinstance(data, dict):
-        base = data
-    else:
-        base = {}
-
-    name = getattr(data, "name", base.get("name", cmd_name)) or cmd_name
-    aliases = getattr(data, "alias", base.get("alias", [])) or []
-    usage = getattr(data, "usage", base.get("usage", ""))
-    summary = getattr(data, "summary", base.get("summary", ""))
-    details = getattr(data, "detail", base.get("detail", [])) or []
-    args_list = getattr(data, "arg", base.get("arg", [])) or []
-    args = [_normalize_arg(arg) for arg in args_list]
-
-    return {
-        "name": str(name).replace("_", "-").lower(),
-        "aliases": [str(a).replace("_", "-").lower() for a in aliases if a],
-        "usage": usage,
-        "summary": summary,
-        "details": details,
-        "args": args,
-        "raw": data,
-    }
-
-
-def list_cmdlet_metadata() -> Dict[str, Dict[str, Any]]:
-    """Collect metadata for all registered cmdlet keyed by canonical name."""
-    ensure_registry_loaded()
-    entries: Dict[str, Dict[str, Any]] = {}
-    for reg_name in (REGISTRY or {}).keys():
-        meta = get_cmdlet_metadata(reg_name)
-        canonical = str(reg_name).replace("_", "-").lower()
-
-        if meta:
-            canonical = meta.get("name", canonical)
-            aliases = meta.get("aliases", [])
-            base = entries.get(
-                canonical,
-                {
-                    "name": canonical,
-                    "aliases": [],
-                    "usage": "",
-                    "summary": "",
-                    "details": [],
-                    "args": [],
-                    "raw": meta.get("raw"),
-                },
-            )
-            merged_aliases = set(base.get("aliases", [])) | set(aliases)
-            if canonical != reg_name:
-                merged_aliases.add(reg_name)
-            base["aliases"] = sorted(a for a in merged_aliases if a and a != canonical)
-            if not base.get("usage") and meta.get("usage"):
-                base["usage"] = meta["usage"]
-            if not base.get("summary") and meta.get("summary"):
-                base["summary"] = meta["summary"]
-            if not base.get("details") and meta.get("details"):
-                base["details"] = meta["details"]
-            if not base.get("args") and meta.get("args"):
-                base["args"] = meta["args"]
-            if not base.get("raw"):
-                base["raw"] = meta.get("raw")
-            entries[canonical] = base
-        else:
-            entries.setdefault(
-                canonical,
-                {"name": canonical, "aliases": [], "usage": "", "summary": "", "details": [], "args": [], "raw": None},
-            )
-    return entries
-
-
-def list_cmdlet_names(include_aliases: bool = True) -> List[str]:
-    """Return sorted cmdlet names (optionally including aliases)."""
-    ensure_registry_loaded()
-    entries = list_cmdlet_metadata()
-    names = set()
-    for meta in entries.values():
-        names.add(meta.get("name", ""))
-        if include_aliases:
-            for alias in meta.get("aliases", []):
-                names.add(alias)
-    return sorted(n for n in names if n)
-
-
-def get_cmdlet_arg_flags(cmd_name: str) -> List[str]:
-    """Return flag variants for cmdlet arguments (e.g., -name/--name)."""
-    meta = get_cmdlet_metadata(cmd_name)
-    if not meta:
-        return []
-
-    # Preserve the order that arguments are defined on the cmdlet (arg=[...]) so
-    # completions feel stable and predictable.
-    flags: List[str] = []
-    seen: set[str] = set()
-
-    for arg in meta.get("args", []):
-        name = str(arg.get("name") or "").strip().lstrip("-")
-        if not name:
-            continue
-        for candidate in (f"-{name}", f"--{name}"):
-            if candidate not in seen:
-                flags.append(candidate)
-                seen.add(candidate)
-
-    return flags
-
-
-def get_cmdlet_arg_choices(cmd_name: str, arg_name: str) -> List[str]:
-    """Return declared choices for a cmdlet argument."""
-    meta = get_cmdlet_metadata(cmd_name)
-    if not meta:
-        return []
-    target = arg_name.lstrip("-")
-    for arg in meta.get("args", []):
-        if arg.get("name") == target:
-            return list(arg.get("choices", []) or [])
-    return []

cmdlet/check_file_status.py

@@ -1,190 +0,0 @@
-from __future__ import annotations
-
-from typing import Any, Dict, Sequence
-import json
-import sys
-
-from SYS.logger import log
-
-from API import HydrusNetwork as hydrus_wrapper
-from ._shared import Cmdlet, CmdletArg, SharedArgs, normalize_hash, should_show_help
-from Store import Store
-
-
-CMDLET = Cmdlet(
-    name="check-file-status",
-    summary="Check if a file is active, deleted, or corrupted in Hydrus.",
-    usage="check-file-status [-hash <sha256>] [-store <name>]",
-    arg=[
-        SharedArgs.HASH,
-        SharedArgs.STORE,
-    ],
-    detail=[
-        "- Shows whether file is active in Hydrus or marked as deleted",
-        "- Detects corrupted data (e.g., comma-separated url)",
-        "- Displays file metadata and service locations",
-        "- Note: Hydrus keeps deleted files for recovery. Use cleanup-corrupted for full removal.",
-    ],
-)
-
-
-def _run(result: Any, args: Sequence[str], config: Dict[str, Any]) -> int:
-    # Help
-    if should_show_help(args):
-        log(f"Cmdlet: {CMDLET.name}\nSummary: {CMDLET.summary}\nUsage: {CMDLET.usage}")
-        return 0
-
-    # Parse arguments
-    override_hash: str | None = None
-    override_store: str | None = None
-    i = 0
-    while i < len(args):
-        token = args[i]
-        low = str(token).lower()
-        if low in {"-hash", "--hash", "hash"} and i + 1 < len(args):
-            override_hash = str(args[i + 1]).strip()
-            i += 2
-            continue
-        if low in {"-store", "--store", "store"} and i + 1 < len(args):
-            override_store = str(args[i + 1]).strip()
-            i += 2
-            continue
-        i += 1
-
-    store_name: str | None = override_store
-    if not store_name:
-        if isinstance(result, dict):
-            store_name = str(result.get("store") or "").strip() or None
-        else:
-            store_name = str(getattr(result, "store", "") or "").strip() or None
-
-    if override_hash:
-        hash_hex = normalize_hash(override_hash)
-    else:
-        if isinstance(result, dict):
-            hash_hex = normalize_hash(result.get("hash") or result.get("hash_hex"))
-        else:
-            hash_hex = normalize_hash(getattr(result, "hash", None) or getattr(result, "hash_hex", None))
-    
-    if not hash_hex:
-        log("No hash provided and no result selected", file=sys.stderr)
-        return 1
-
-    try:
-        client = None
-        if store_name:
-            # Store specified: do not fall back to a global/default Hydrus client.
-            try:
-                store = Store(config)
-                backend = store[str(store_name)]
-                candidate = getattr(backend, "_client", None)
-                if candidate is not None and hasattr(candidate, "fetch_file_metadata"):
-                    client = candidate
-            except Exception:
-                client = None
-
-            if client is None:
-                log(f"Hydrus client unavailable for store '{store_name}'", file=sys.stderr)
-                return 1
-        else:
-            client = hydrus_wrapper.get_client(config)
-
-        if client is None:
-            log("Hydrus client unavailable", file=sys.stderr)
-            return 1
-    except Exception as exc:
-        log(f"Hydrus client unavailable: {exc}", file=sys.stderr)
-        return 1
-
-    try:
-        result_data = client.fetch_file_metadata(hashes=[hash_hex])
-        if not result_data.get("metadata"):
-            log(f"File not found: {hash_hex[:16]}...", file=sys.stderr)
-            return 1
-
-        file_info = result_data["metadata"][0]
-        
-        # Status summary
-        is_deleted = file_info.get("is_deleted", False)
-        is_local = file_info.get("is_local", False)
-        is_trashed = file_info.get("is_trashed", False)
-        
-        status_str = "DELETED" if is_deleted else ("TRASHED" if is_trashed else "ACTIVE")
-        log(f"File status: {status_str}", file=sys.stderr)
-        
-        # File info
-        log(f"\n📄 File Information:", file=sys.stderr)
-        log(f"  Hash: {file_info['hash'][:16]}...", file=sys.stderr)
-        log(f"  Size: {file_info['size']:,} bytes", file=sys.stderr)
-        log(f"  MIME: {file_info['mime']}", file=sys.stderr)
-        log(f"  Dimensions: {file_info.get('width', '?')}x{file_info.get('height', '?')}", file=sys.stderr)
-        
-        # Service status
-        file_services = file_info.get("file_services", {})
-        current_services = file_services.get("current", {})
-        deleted_services = file_services.get("deleted", {})
-        
-        if current_services:
-            log(f"\n✓ In services ({len(current_services)}):", file=sys.stderr)
-            for service_key, service_info in current_services.items():
-                sname = service_info.get("name", "unknown")
-                stype = service_info.get("type_pretty", "unknown")
-                log(f"    - {sname} ({stype})", file=sys.stderr)
-        
-        if deleted_services:
-            log(f"\n✗ Deleted from services ({len(deleted_services)}):", file=sys.stderr)
-            for service_key, service_info in deleted_services.items():
-                sname = service_info.get("name", "unknown")
-                stype = service_info.get("type_pretty", "unknown")
-                time_deleted = service_info.get("time_deleted", "?")
-                log(f"    - {sname} ({stype}) - deleted at {time_deleted}", file=sys.stderr)
-        
-        # URL check
-        url = file_info.get("url", [])
-        log(f"\n🔗 url ({len(url)}):", file=sys.stderr)
-        
-        corrupted_count = 0
-        for i, url in enumerate(url, 1):
-            if "," in url:
-                corrupted_count += 1
-                log(f"    [{i}] ⚠️ CORRUPTED (comma-separated): {url[:50]}...", file=sys.stderr)
-            else:
-                log(f"    [{i}] {url[:70]}{'...' if len(url) > 70 else ''}", file=sys.stderr)
-        
-        if corrupted_count > 0:
-            log(f"\n⚠️ WARNING: Found {corrupted_count} corrupted URL(s)", file=sys.stderr)
-        
-        # Tags
-        tags_dict = file_info.get("tags", {})
-        total_tags = 0
-        for service_key, service_data in tags_dict.items():
-            service_name = service_data.get("name", "unknown")
-            display_tags = service_data.get("display_tags", {}).get("0", [])
-            total_tags += len(display_tags)
-        
-        if total_tags > 0:
-            log(f"\n🏷️ Tags ({total_tags}):", file=sys.stderr)
-            for service_key, service_data in tags_dict.items():
-                display_tags = service_data.get("display_tags", {}).get("0", [])
-                if display_tags:
-                    service_name = service_data.get("name", "unknown")
-                    log(f"    {service_name}:", file=sys.stderr)
-                    for tag in display_tags[:5]:  # Show first 5
-                        log(f"      - {tag}", file=sys.stderr)
-                    if len(display_tags) > 5:
-                        log(f"      ... and {len(display_tags) - 5} more", file=sys.stderr)
-        
-        log("\n", file=sys.stderr)
-        return 0
-        
-    except Exception as exc:
-        log(f"Error checking file status: {exc}", file=sys.stderr)
-        import traceback
-        traceback.print_exc(file=sys.stderr)
-        return 1
-
-
-# Register cmdlet (no legacy decorator)
-CMDLET.exec = _run
-CMDLET.alias = ["check-status", "file-status", "status"]
-CMDLET.register()

cmdlet/cleanup.py

@@ -1,105 +0,0 @@
-"""Cleanup cmdlet for removing temporary artifacts from pipeline.
-
-This cmdlet processes result lists and removes temporary files (marked with is_temp=True),
-then emits the remaining non-temporary results for further pipeline stages.
-"""
-
-from __future__ import annotations
-
-from typing import Any, Dict, Sequence
-from pathlib import Path
-import sys
-import json
-
-from SYS.logger import log
-
-from ._shared import Cmdlet, CmdletArg, get_pipe_object_path, normalize_result_input, filter_results_by_temp, should_show_help
-import models
-import pipeline as pipeline_context
-
-def _run(result: Any, args: Sequence[str], config: Dict[str, Any]) -> int:
-    """Remove temporary files from pipeline results.
-    
-    Accepts:
-    - Single result object with is_temp field
-    - List of result objects to clean up
-    
-    Process:
-    - Filters results by is_temp=True
-    - Deletes those files from disk
-    - Emits only non-temporary results
-    
-    Typical pipeline usage:
-    download-data url | screen-shot | add-tag -store local "tag" --all | cleanup
-    """
-    
-    # Help
-    if should_show_help(args):
-        log(f"Cmdlet: {CMDLET.name}\nSummary: {CMDLET.summary}\nUsage: {CMDLET.usage}")
-        return 0
-    
-    # Normalize input to list
-    results = normalize_result_input(result)
-    
-    if not results:
-        log("[cleanup] No results to process", file=sys.stderr)
-        return 1
-    
-    # Separate temporary and permanent results
-    temp_results = pipeline_context.filter_results_by_temp(results, include_temp=True)
-    perm_results = pipeline_context.filter_results_by_temp(results, include_temp=False)
-    
-    # Delete temporary files
-    deleted_count = 0
-    for temp_result in temp_results:
-        try:
-            file_path = get_pipe_object_path(temp_result)
-            
-            if file_path:
-                path_obj = Path(file_path)
-                if path_obj.exists():
-                    # Delete the file
-                    path_obj.unlink()
-                    log(f"[cleanup] Deleted temporary file: {path_obj.name}", file=sys.stderr)
-                    deleted_count += 1
-                    
-                    # Clean up any associated sidecar files
-                    for ext in ['.tag', '.metadata']:
-                        sidecar = path_obj.parent / (path_obj.name + ext)
-                        if sidecar.exists():
-                            try:
-                                sidecar.unlink()
-                                log(f"[cleanup] Deleted sidecar: {sidecar.name}", file=sys.stderr)
-                            except Exception as e:
-                                log(f"[cleanup] Warning: Could not delete sidecar {sidecar.name}: {e}", file=sys.stderr)
-                else:
-                    log(f"[cleanup] File does not exist: {file_path}", file=sys.stderr)
-        except Exception as e:
-            log(f"[cleanup] Error deleting file: {e}", file=sys.stderr)
-    
-    # Log summary
-    log(f"[cleanup] Deleted {deleted_count} temporary file(s), emitting {len(perm_results)} permanent result(s)", file=sys.stderr)
-    
-    # Emit permanent results for downstream processing
-    for perm_result in perm_results:
-        pipeline_context.emit(perm_result)
-    
-    return 0
-
-
-CMDLET = Cmdlet(
-    name="cleanup",
-    summary="Remove temporary artifacts from pipeline (marked with is_temp=True).",
-    usage="cleanup",
-    arg=[],
-    detail=[
-        "- Accepts pipeline results that may contain temporary files (screenshots, intermediate artifacts)",
-        "- Deletes files marked with is_temp=True from disk",
-        "- Also cleans up associated sidecar files (.tag, .metadata)",
-        "- Emits only non-temporary results for further processing",
-        "- Typical usage at end of pipeline: ... | add-tag -store local \"tag\" --all | cleanup",
-        "- Exit code 0 if cleanup successful, 1 if no results to process",
-    ],
-    exec=_run,
-).register()
-

cmdlet/delete_file.py

@@ -8,12 +8,12 @@ from pathlib import Path
 from SYS.logger import debug, log
 from Store.Folder import Folder
 from Store import Store
-from ._shared import Cmdlet, CmdletArg, normalize_hash, looks_like_hash, get_field, should_show_help
+from . import _shared as sh
 from API import HydrusNetwork as hydrus_wrapper
 import pipeline as ctx
 
 
-class Delete_File(Cmdlet):
+class Delete_File(sh.Cmdlet):
     """Class-based delete-file cmdlet with self-registration."""
 
     def __init__(self) -> None:
@@ -23,10 +23,10 @@ class Delete_File(Cmdlet):
             usage="delete-file [-hash <sha256>] [-conserve <local|hydrus>] [-lib-root <path>] [reason]",
             alias=["del-file"],
             arg=[
-                CmdletArg("hash", description="Override the Hydrus file hash (SHA256) to target instead of the selected result."),
-                CmdletArg("conserve", description="Choose which copy to keep: 'local' or 'hydrus'."),
-                CmdletArg("lib-root", description="Path to local library root for database cleanup."),
-                CmdletArg("reason", description="Optional reason for deletion (free text)."),
+                sh.CmdletArg("hash", description="Override the Hydrus file hash (SHA256) to target instead of the selected result."),
+                sh.CmdletArg("conserve", description="Choose which copy to keep: 'local' or 'hydrus'."),
+                sh.CmdletArg("lib-root", description="Path to local library root for database cleanup."),
+                sh.CmdletArg("reason", description="Optional reason for deletion (free text)."),
             ],
             detail=[
                 "Default removes both the local file and Hydrus file.",
@@ -45,24 +45,28 @@ class Delete_File(Cmdlet):
         if isinstance(item, dict):
             hash_hex_raw = item.get("hash_hex") or item.get("hash")
             target = item.get("target") or item.get("file_path") or item.get("path")
+            title_val = item.get("title") or item.get("name")
         else:
-            hash_hex_raw = get_field(item, "hash_hex") or get_field(item, "hash")
-            target = get_field(item, "target") or get_field(item, "file_path") or get_field(item, "path")
+            hash_hex_raw = sh.get_field(item, "hash_hex") or sh.get_field(item, "hash")
+            target = sh.get_field(item, "target") or sh.get_field(item, "file_path") or sh.get_field(item, "path")
+            title_val = sh.get_field(item, "title") or sh.get_field(item, "name")
         
         store = None
         if isinstance(item, dict):
             store = item.get("store")
         else:
-            store = get_field(item, "store")
+            store = sh.get_field(item, "store")
 
         store_lower = str(store).lower() if store else ""
         is_hydrus_store = bool(store_lower) and ("hydrus" in store_lower or store_lower in {"home", "work"})
+        store_label = str(store) if store else "default"
+        hydrus_prefix = f"[hydrusnetwork:{store_label}]"
         
         # For Hydrus files, the target IS the hash
         if is_hydrus_store and not hash_hex_raw:
             hash_hex_raw = target
 
-        hash_hex = normalize_hash(override_hash) if override_hash else normalize_hash(hash_hex_raw)
+        hash_hex = sh.normalize_hash(override_hash) if override_hash else sh.normalize_hash(hash_hex_raw)
 
         local_deleted = False
         local_target = isinstance(target, str) and target.strip() and not str(target).lower().startswith(("http://", "https://"))
@@ -156,19 +160,28 @@ class Delete_File(Cmdlet):
             try:
                 client._post("/add_files/delete_files", data=payload)  # type: ignore[attr-defined]
                 hydrus_deleted = True
-                preview = hash_hex[:12] + ('…' if len(hash_hex) > 12 else '')
-                debug(f"Deleted from Hydrus: {preview}…", file=sys.stderr)
+                title_str = str(title_val).strip() if title_val else ""
+                if title_str:
+                    debug(f"{hydrus_prefix} Deleted title:{title_str} hash:{hash_hex}", file=sys.stderr)
+                else:
+                    debug(f"{hydrus_prefix} Deleted hash:{hash_hex}", file=sys.stderr)
             except Exception:
                 # If it's not in Hydrus (e.g. 404 or similar), that's fine
                 if not local_deleted:
                     return False
 
         if hydrus_deleted and hash_hex:
-            preview = hash_hex[:12] + ('…' if len(hash_hex) > 12 else '')
+            title_str = str(title_val).strip() if title_val else ""
             if reason:
-                ctx.emit(f"Deleted {preview} (reason: {reason}).")
+                if title_str:
+                    ctx.emit(f"{hydrus_prefix} Deleted title:{title_str} hash:{hash_hex} (reason: {reason}).")
+                else:
+                    ctx.emit(f"{hydrus_prefix} Deleted hash:{hash_hex} (reason: {reason}).")
             else:
-                ctx.emit(f"Deleted {preview}.")
+                if title_str:
+                    ctx.emit(f"{hydrus_prefix} Deleted title:{title_str} hash:{hash_hex}.")
+                else:
+                    ctx.emit(f"{hydrus_prefix} Deleted hash:{hash_hex}.")
 
         if hydrus_deleted or local_deleted:
             return True
@@ -178,7 +191,7 @@ class Delete_File(Cmdlet):
 
     def run(self, result: Any, args: Sequence[str], config: Dict[str, Any]) -> int:
         """Execute delete-file command."""
-        if should_show_help(args):
+        if sh.should_show_help(args):
             log(f"Cmdlet: {self.name}\nSummary: {self.summary}\nUsage: {self.usage}")
             return 0
 

cmdlet/delete_note.py

@@ -7,16 +7,16 @@ import sys
 from SYS.logger import log
 
 import pipeline as ctx
-from ._shared import (
-    Cmdlet,
-    CmdletArg,
-    SharedArgs,
-    normalize_hash,
-    parse_cmdlet_args,
-    normalize_result_input,
-    get_field,
-    should_show_help,
-)
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+normalize_hash = sh.normalize_hash
+parse_cmdlet_args = sh.parse_cmdlet_args
+normalize_result_input = sh.normalize_result_input
+get_field = sh.get_field
+should_show_help = sh.should_show_help
 from Store import Store
 from SYS.utils import sha256_file
 

cmdlet/delete_relationship.py

@@ -10,7 +10,16 @@ import sys
 from SYS.logger import log
 
 import pipeline as ctx
-from ._shared import Cmdlet, CmdletArg, SharedArgs, parse_cmdlet_args, normalize_hash, normalize_result_input, get_field, should_show_help
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+parse_cmdlet_args = sh.parse_cmdlet_args
+normalize_hash = sh.normalize_hash
+normalize_result_input = sh.normalize_result_input
+get_field = sh.get_field
+should_show_help = sh.should_show_help
 from API.folder import API_folder_store
 from Store import Store
 from config import get_local_storage_path

cmdlet/delete_tag.py

@@ -7,7 +7,15 @@ import sys
 
 import models
 import pipeline as ctx
-from ._shared import Cmdlet, CmdletArg, SharedArgs, normalize_hash, parse_tag_arguments, should_show_help, get_field
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+normalize_hash = sh.normalize_hash
+parse_tag_arguments = sh.parse_tag_arguments
+should_show_help = sh.should_show_help
+get_field = sh.get_field
 from SYS.logger import debug, log
 from Store import Store
 

cmdlet/delete_url.py

@@ -4,7 +4,16 @@ from typing import Any, Dict, Sequence
 import sys
 
 import pipeline as ctx
-from ._shared import Cmdlet, CmdletArg, SharedArgs, parse_cmdlet_args, get_field, normalize_hash
+from . import _shared as sh
+
+Cmdlet, CmdletArg, SharedArgs, parse_cmdlet_args, get_field, normalize_hash = (
+    sh.Cmdlet,
+    sh.CmdletArg,
+    sh.SharedArgs,
+    sh.parse_cmdlet_args,
+    sh.get_field,
+    sh.normalize_hash,
+)
 from SYS.logger import log
 from Store import Store
 

cmdlet/download_file.py

@@ -17,15 +17,15 @@ from SYS.download import DownloadError, _download_direct_file
 from SYS.logger import log, debug
 import pipeline as pipeline_context
 
-from ._shared import (
-    Cmdlet,
-    CmdletArg,
-    SharedArgs,
-    parse_cmdlet_args,
-    register_url_with_local_library,
-    coerce_to_pipe_object,
-    get_field,
-)
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+parse_cmdlet_args = sh.parse_cmdlet_args
+register_url_with_local_library = sh.register_url_with_local_library
+coerce_to_pipe_object = sh.coerce_to_pipe_object
+get_field = sh.get_field
 
 
 class Download_File(Cmdlet):
@@ -251,6 +251,13 @@ class Download_File(Cmdlet):
 
                     # Fallback: if we have a direct HTTP URL, download it directly
                     if downloaded_path is None and isinstance(target, str) and target.startswith("http"):
+                        # Guard: provider landing pages (e.g. LibGen ads.php) are HTML, not files.
+                        # Never download these as "files".
+                        if str(table or "").lower() == "libgen":
+                            low = target.lower()
+                            if ("/ads.php" in low) or ("/file.php" in low) or ("/index.php" in low):
+                                log("[download-file] Refusing to download LibGen landing page (expected provider to resolve file link)", file=sys.stderr)
+                                continue
                         debug(f"[download-file] Provider item looks like direct URL, downloading: {target}")
                         result_obj = _download_direct_file(target, final_output_dir, quiet=quiet_mode)
                         file_path = None

cmdlet/download_media.py

@@ -38,7 +38,18 @@ from models import DownloadError, DownloadOptions, DownloadMediaResult, DebugLog
 import pipeline as pipeline_context
 from result_table import ResultTable
 
-from ._shared import Cmdlet, CmdletArg, SharedArgs, create_pipe_object_result, parse_cmdlet_args, register_url_with_local_library, coerce_to_pipe_object
+from tool.ytdlp import YtDlpTool
+
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+create_pipe_object_result = sh.create_pipe_object_result
+parse_cmdlet_args = sh.parse_cmdlet_args
+register_url_with_local_library = sh.register_url_with_local_library
+coerce_to_pipe_object = sh.coerce_to_pipe_object
+get_field = sh.get_field
 
 
 # Minimal inlined helpers from helper/download.py (is_url_supported_by_ytdlp, list_formats)
@@ -62,6 +73,136 @@ _EXTRACTOR_CACHE: List[Any] | None = None
 _YTDLP_PROGRESS_BAR = ProgressBar()
 
 
+_SUBTITLE_EXTS = (".vtt", ".srt", ".ass", ".ssa", ".lrc")
+
+
+def _format_chapters_note(info: Dict[str, Any]) -> Optional[str]:
+    """Format yt-dlp chapter metadata into a stable, note-friendly text.
+
+    Output is one chapter per line, e.g.:
+      00:00 Intro
+      01:23-02:10 Topic name
+    """
+    try:
+        chapters = info.get("chapters")
+    except Exception:
+        chapters = None
+
+    if not isinstance(chapters, list) or not chapters:
+        return None
+
+    rows: List[tuple[int, Optional[int], str]] = []
+    max_t = 0
+    for ch in chapters:
+        if not isinstance(ch, dict):
+            continue
+        start_raw = ch.get("start_time")
+        end_raw = ch.get("end_time")
+        title_raw = ch.get("title") or ch.get("name") or ch.get("chapter")
+
+        try:
+            start_s = int(float(start_raw))
+        except Exception:
+            continue
+
+        end_s: Optional[int] = None
+        try:
+            if end_raw is not None:
+                end_s = int(float(end_raw))
+        except Exception:
+            end_s = None
+
+        title = str(title_raw).strip() if title_raw is not None else ""
+        rows.append((start_s, end_s, title))
+        try:
+            max_t = max(max_t, start_s, end_s or 0)
+        except Exception:
+            max_t = max(max_t, start_s)
+
+    if not rows:
+        return None
+
+    force_hours = bool(max_t >= 3600)
+
+    def _tc(seconds: int) -> str:
+        total = max(0, int(seconds))
+        minutes, secs = divmod(total, 60)
+        hours, minutes = divmod(minutes, 60)
+        if force_hours:
+            return f"{hours:02d}:{minutes:02d}:{secs:02d}"
+        return f"{minutes:02d}:{secs:02d}"
+
+    lines: List[str] = []
+    for start_s, end_s, title in sorted(rows, key=lambda r: (r[0], r[1] if r[1] is not None else 10**9, r[2])):
+        if end_s is not None and end_s > start_s:
+            prefix = f"{_tc(start_s)}-{_tc(end_s)}"
+        else:
+            prefix = _tc(start_s)
+        line = f"{prefix} {title}".strip()
+        if line:
+            lines.append(line)
+
+    text = "\n".join(lines).strip()
+    return text or None
+
+
+def _best_subtitle_sidecar(media_path: Path) -> Optional[Path]:
+    """Find the most likely subtitle sidecar file for a downloaded media file."""
+    try:
+        base_dir = media_path.parent
+        stem = media_path.stem
+        if not stem:
+            return None
+
+        candidates: List[Path] = []
+        for p in base_dir.glob(stem + ".*"):
+            try:
+                if not p.is_file():
+                    continue
+            except Exception:
+                continue
+            if p.suffix.lower() in _SUBTITLE_EXTS:
+                candidates.append(p)
+
+        if not candidates:
+            return None
+
+        def _rank(path: Path) -> tuple[int, int, float, str]:
+            name = path.name.lower()
+            lang_rank = 0 if ".en." in name or name.endswith(".en" + path.suffix.lower()) else 1
+            ext = path.suffix.lower()
+            ext_rank_map = {".vtt": 0, ".srt": 1, ".ass": 2, ".ssa": 3, ".lrc": 4}
+            ext_rank = ext_rank_map.get(ext, 9)
+            try:
+                mtime = float(path.stat().st_mtime)
+            except Exception:
+                mtime = 0.0
+            return (lang_rank, ext_rank, -mtime, name)
+
+        candidates.sort(key=_rank)
+        return candidates[0]
+    except Exception:
+        return None
+
+
+def _read_text_file(path: Path, *, max_bytes: int = 1_500_000) -> Optional[str]:
+    try:
+        data = path.read_bytes()
+    except Exception:
+        return None
+    if not data:
+        return None
+    if len(data) > max_bytes:
+        data = data[:max_bytes]
+    try:
+        return data.decode("utf-8", errors="replace")
+    except Exception:
+        try:
+            return data.decode(errors="replace")
+        except Exception:
+            return None
+
+
 def _ensure_yt_dlp_ready() -> None:
     if yt_dlp is not None:
         return
@@ -100,16 +241,26 @@ def list_formats(url: str, no_playlist: bool = False, playlist_items: Optional[s
             ydl_opts["noplaylist"] = True
         if playlist_items:
             ydl_opts["playlist_items"] = playlist_items
+
+        debug(f"Fetching format list for: {url}")
         with yt_dlp.YoutubeDL(ydl_opts) as ydl:  # type: ignore[arg-type]
-            debug(f"Fetching format list for: {url}")
             info = ydl.extract_info(url, download=False)
-            formats = info.get("formats", [])
-            if not formats:
-                log("No formats available", file=sys.stderr)
-                return None
-            result_formats = []
-            for fmt in formats:
-                result_formats.append({
+
+        if not isinstance(info, dict):
+            log("No formats available", file=sys.stderr)
+            return None
+
+        formats = info.get("formats") or []
+        if not isinstance(formats, list) or not formats:
+            log("No formats available", file=sys.stderr)
+            return None
+
+        result_formats: List[Dict[str, Any]] = []
+        for fmt in formats:
+            if not isinstance(fmt, dict):
+                continue
+            result_formats.append(
+                {
                     "format_id": fmt.get("format_id", ""),
                     "format": fmt.get("format", ""),
                     "ext": fmt.get("ext", ""),
@@ -122,9 +273,11 @@ def list_formats(url: str, no_playlist: bool = False, playlist_items: Optional[s
                     "filesize": fmt.get("filesize"),
                     "abr": fmt.get("abr"),
                     "tbr": fmt.get("tbr"),
-                })
-            debug(f"Found {len(result_formats)} available formats")
-            return result_formats
+                }
+            )
+
+        debug(f"Found {len(result_formats)} available formats")
+        return result_formats or None
     except Exception as e:
         log(f"✗ Error fetching formats: {e}", file=sys.stderr)
         return None
@@ -215,6 +368,31 @@ def _download_with_sections_via_cli(url: str, ytdl_options: Dict[str, Any], sect
         cmd = ["yt-dlp"]
         if ytdl_options.get("format"):
             cmd.extend(["-f", ytdl_options["format"]])
+        if ytdl_options.get("merge_output_format"):
+            cmd.extend(["--merge-output-format", str(ytdl_options["merge_output_format"])])
+
+        # For CLI downloads, infer chapter/metadata embedding from either legacy flags
+        # or explicit FFmpegMetadata postprocessor entries.
+        postprocessors = ytdl_options.get("postprocessors")
+        want_add_metadata = bool(ytdl_options.get("addmetadata"))
+        want_embed_chapters = bool(ytdl_options.get("embedchapters"))
+        if isinstance(postprocessors, list):
+            for pp in postprocessors:
+                if not isinstance(pp, dict):
+                    continue
+                if str(pp.get("key") or "") == "FFmpegMetadata":
+                    want_add_metadata = True
+                    if bool(pp.get("add_chapters", True)):
+                        want_embed_chapters = True
+
+        if want_add_metadata:
+            cmd.append("--add-metadata")
+        if want_embed_chapters:
+            cmd.append("--embed-chapters")
+        if ytdl_options.get("writesubtitles"):
+            cmd.append("--write-sub")
+            cmd.append("--write-auto-sub")
+            cmd.extend(["--sub-format", "vtt"])
         if ytdl_options.get("force_keyframes_at_cuts"):
             cmd.extend(["--force-keyframes-at-cuts"]) if ytdl_options.get("force_keyframes_at_cuts") else None
         cmd.extend(["-o", section_outtmpl])
@@ -258,11 +436,6 @@ def _build_ytdlp_options(opts: DownloadOptions) -> Dict[str, Any]:
 
     if opts.cookies_path and opts.cookies_path.is_file():
         base_options["cookiefile"] = str(opts.cookies_path)
-    else:
-        from hydrus_health_check import get_cookies_file_path  # local import
-        global_cookies = get_cookies_file_path()
-        if global_cookies:
-            base_options["cookiefile"] = global_cookies
 
     if opts.no_playlist:
         base_options["noplaylist"] = True
@@ -274,6 +447,37 @@ def _build_ytdlp_options(opts: DownloadOptions) -> Dict[str, Any]:
         base_options["format"] = opts.ytdl_format or "bestvideo+bestaudio/best"
         base_options["format_sort"] = ["res:4320", "res:2880", "res:2160", "res:1440", "res:1080", "res:720", "res"]
 
+    # Optional yt-dlp features
+    if getattr(opts, "embed_chapters", False):
+        # Prefer explicit FFmpegMetadata PP so chapter embedding runs even when
+        # we already specified other postprocessors (e.g. FFmpegExtractAudio).
+        pps = base_options.get("postprocessors")
+        if not isinstance(pps, list):
+            pps = []
+        already_has_metadata = any(
+            isinstance(pp, dict) and str(pp.get("key") or "") == "FFmpegMetadata" for pp in pps
+        )
+        if not already_has_metadata:
+            pps.append(
+                {
+                    "key": "FFmpegMetadata",
+                    "add_metadata": True,
+                    "add_chapters": True,
+                    "add_infojson": "if_exists",
+                }
+            )
+        base_options["postprocessors"] = pps
+
+        # Chapter embedding is most reliable in mkv/mp4 containers.
+        # When merging separate video+audio streams, prefer mkv so mpv sees chapters.
+        if opts.mode != "audio":
+            base_options.setdefault("merge_output_format", "mkv")
+
+    if getattr(opts, "write_sub", False):
+        base_options["writesubtitles"] = True
+        base_options["writeautomaticsub"] = True
+        base_options["subtitlesformat"] = "vtt"
+
     if opts.clip_sections:
         sections: List[str] = []
 
@@ -410,13 +614,27 @@ def _get_libgen_download_url(libgen_url: str) -> Optional[str]:
             response = session.get(libgen_url, timeout=10, allow_redirects=True)
             final_url = response.url
             try:
-                from bs4 import BeautifulSoup
-                soup = BeautifulSoup(response.content, 'html.parser')
-                for link in soup.find_all('a'):
-                    href = link.get('href')
-                    if href and 'get.php' in href:
-                        return urljoin(libgen_url, href)
-            except ImportError:
+                try:
+                    from lxml import html as lxml_html
+                except ImportError:
+                    lxml_html = None
+
+                if lxml_html is not None:
+                    doc = lxml_html.fromstring(response.content)
+                    for a in doc.xpath("//a[@href]"):
+                        href = str(a.get("href") or "").strip()
+                        if href and "get.php" in href.lower():
+                            return urljoin(final_url, href)
+                else:
+                    for m in re.finditer(
+                        r"href=[\"\']([^\"\']+)[\"\']",
+                        response.text or "",
+                        flags=re.IGNORECASE,
+                    ):
+                        href = str(m.group(1) or "").strip()
+                        if href and "get.php" in href.lower():
+                            return urljoin(final_url, href)
+            except Exception:
                 pass
             if final_url != libgen_url:
                 debug(f"LibGen resolved to mirror: {final_url}")
@@ -648,7 +866,7 @@ def _download_direct_file(
         raise DownloadError(f"Error downloading file: {exc}") from exc
 
 
-def probe_url(url: str, no_playlist: bool = False, timeout_seconds: int = 15) -> Optional[Dict[str, Any]]:
+def probe_url(url: str, no_playlist: bool = False, timeout_seconds: int = 15, *, cookiefile: Optional[str] = None) -> Optional[Dict[str, Any]]:
     """Probe URL to extract metadata WITHOUT downloading.
     
     Args:
@@ -686,12 +904,8 @@ def probe_url(url: str, no_playlist: bool = False, timeout_seconds: int = 15) ->
                 "noprogress": True,  # No progress bars
             }
             
-            # Add cookies if available (lazy import to avoid circular dependency)
-            from hydrus_health_check import get_cookies_file_path  # local import
-
-            global_cookies = get_cookies_file_path()
-            if global_cookies:
-                ydl_opts["cookiefile"] = global_cookies
+            if cookiefile:
+                ydl_opts["cookiefile"] = str(cookiefile)
             
             # Add no_playlist option if specified
             if no_playlist:
@@ -807,7 +1021,14 @@ def download_media(
             debug(f"Skipping probe for playlist (item selection: {opts.playlist_items}), proceeding with download")
             probe_result = {"url": opts.url}  # Minimal probe result
         else:
-            probe_result = probe_url(opts.url, no_playlist=opts.no_playlist, timeout_seconds=15)
+            probe_cookiefile = None
+            try:
+                if opts.cookies_path and opts.cookies_path.is_file():
+                    probe_cookiefile = str(opts.cookies_path)
+            except Exception:
+                probe_cookiefile = None
+
+            probe_result = probe_url(opts.url, no_playlist=opts.no_playlist, timeout_seconds=15, cookiefile=probe_cookiefile)
         
         if probe_result is None:
             if not opts.quiet:
@@ -1182,6 +1403,8 @@ class Download_Media(Cmdlet):
         try:
             debug("Starting download-media")
 
+            ytdlp_tool = YtDlpTool(config)
+
             # Parse arguments
             parsed = parse_cmdlet_args(args, self)
 
@@ -1192,7 +1415,6 @@ class Download_Media(Cmdlet):
             
             # If no url provided via args, try to extract from piped result
             if not raw_url and result:
-                from ._shared import get_field
                 # Handle single result or list of results
                 results_to_check = result if isinstance(result, list) else [result]
                 for item in results_to_check:
@@ -1226,6 +1448,10 @@ class Download_Media(Cmdlet):
             # Get other options
             clip_spec = parsed.get("clip")
 
+            # Always enable chapters + subtitles so downstream pipes (e.g. mpv) can consume them.
+            embed_chapters = True
+            write_sub = True
+
             mode = "audio" if parsed.get("audio") else "video"
 
             # Parse clip range(s) if specified
@@ -1379,7 +1605,14 @@ class Download_Media(Cmdlet):
                 if playlist_items:
                     return str(requested_url)
                 try:
-                    pr = probe_url(requested_url, no_playlist=False, timeout_seconds=15)
+                    cf = None
+                    try:
+                        cookie_path = ytdlp_tool.resolve_cookiefile()
+                        if cookie_path is not None and cookie_path.is_file():
+                            cf = str(cookie_path)
+                    except Exception:
+                        cf = None
+                    pr = probe_url(requested_url, no_playlist=False, timeout_seconds=15, cookiefile=cf)
                     if isinstance(pr, dict):
                         for key in ("webpage_url", "original_url", "url", "requested_url"):
                             value = pr.get(key)
@@ -1458,7 +1691,14 @@ class Download_Media(Cmdlet):
                         - selected_urls: Optional[List[str]] (expanded per-entry urls when available)
                 """
                 try:
-                    pr = probe_url(url, no_playlist=False, timeout_seconds=15)
+                    cf = None
+                    try:
+                        cookie_path = ytdlp_tool.resolve_cookiefile()
+                        if cookie_path is not None and cookie_path.is_file():
+                            cf = str(cookie_path)
+                    except Exception:
+                        cf = None
+                    pr = probe_url(url, no_playlist=False, timeout_seconds=15, cookiefile=cf)
                 except Exception:
                     pr = None
                 if not isinstance(pr, dict):
@@ -1685,6 +1925,15 @@ class Download_Media(Cmdlet):
                         acodec = fmt.get("acodec", "none")
                         filesize = fmt.get("filesize")
                         format_id = fmt.get("format_id", "")
+
+                        # If the chosen format is video-only (no audio stream), automatically
+                        # request best audio too so the resulting file has sound.
+                        selection_format_id = format_id
+                        try:
+                            if vcodec != "none" and acodec == "none" and format_id:
+                                selection_format_id = f"{format_id}+ba"
+                        except Exception:
+                            selection_format_id = format_id
                         
                         # Format size
                         size_str = ""
@@ -1729,9 +1978,9 @@ class Download_Media(Cmdlet):
                             "full_metadata": {
                                 "format_id": format_id,
                                 "url": url,
-                                "item_selector": format_id,
+                                "item_selector": selection_format_id,
                             },
-                            "_selection_args": ["-format", format_id]
+                            "_selection_args": ["-format", selection_format_id]
                         }
                         
                         # Add to results list and table (don't emit - formats should wait for @N selection)
@@ -1778,23 +2027,57 @@ class Download_Media(Cmdlet):
                             actual_format = playlist_items
                             actual_playlist_items = None
 
-                    # Auto-pick best audio format when -audio is used and no explicit format is given.
+                    # For -audio, default to yt-dlp's built-in bestaudio selector.
+                    # This should *not* require interactive format picking.
                     if mode == "audio" and not actual_format:
-                        chosen = None
-                        formats = list_formats(url, no_playlist=False, playlist_items=actual_playlist_items)
-                        if formats:
-                            chosen = _pick_best_audio_format_id(formats)
-                        actual_format = chosen or "bestaudio/best"
+                        actual_format = "bestaudio"
+
+                    # If no explicit format is provided for video mode, allow a config override.
+                    if mode == "video" and not actual_format:
+                        configured = (ytdlp_tool.default_format("video") or "").strip()
+                        if configured and configured != "bestvideo+bestaudio/best":
+                            actual_format = configured
+
+                    # If a single format id was chosen and it is video-only, auto-merge best audio.
+                    if (
+                        actual_format
+                        and isinstance(actual_format, str)
+                        and mode != "audio"
+                        and "+" not in actual_format
+                        and "/" not in actual_format
+                        and "[" not in actual_format
+                        and actual_format not in {"best", "bv", "ba", "b"}
+                    ):
+                        try:
+                            formats = list_formats(url, no_playlist=False, playlist_items=actual_playlist_items)
+                            if formats:
+                                fmt_match = next(
+                                    (f for f in formats if str(f.get("format_id", "")) == actual_format),
+                                    None,
+                                )
+                                if fmt_match:
+                                    vcodec = str(fmt_match.get("vcodec", "none"))
+                                    acodec = str(fmt_match.get("acodec", "none"))
+                                    if vcodec != "none" and acodec == "none":
+                                        debug(
+                                            f"Selected video-only format {actual_format}; using {actual_format}+ba for audio"
+                                        )
+                                        actual_format = f"{actual_format}+ba"
+                        except Exception:
+                            pass
 
                     opts = DownloadOptions(
                         url=url,
                         mode=mode,
                         output_dir=final_output_dir,
                         ytdl_format=actual_format,
+                        cookies_path=ytdlp_tool.resolve_cookiefile(),
                         clip_sections=clip_sections_spec,
                         playlist_items=actual_playlist_items,
                         quiet=quiet_mode,
                         no_playlist=False,
+                        embed_chapters=embed_chapters,
+                        write_sub=write_sub,
                     )
 
                     # Use timeout wrapper to prevent hanging
@@ -1838,7 +2121,40 @@ class Download_Media(Cmdlet):
                     # Build PipeObjects first so we can attach cross-clip relationships.
                     pipe_objects: List[Dict[str, Any]] = []
                     for downloaded in results_to_emit:
-                        pipe_objects.append(self._build_pipe_object(downloaded, url, opts))
+                        po = self._build_pipe_object(downloaded, url, opts)
+
+                        # Attach chapter timestamps for downstream consumers (e.g., mpv scripts)
+                        # even if container embedding fails.
+                        try:
+                            info = downloaded.info if isinstance(getattr(downloaded, "info", None), dict) else {}
+                        except Exception:
+                            info = {}
+                        chapters_text = _format_chapters_note(info) if embed_chapters else None
+                        if chapters_text:
+                            notes = po.get("notes")
+                            if not isinstance(notes, dict):
+                                notes = {}
+                            notes.setdefault("chapters", chapters_text)
+                            po["notes"] = notes
+
+                        if write_sub:
+                            try:
+                                media_path = Path(str(po.get("path") or ""))
+                            except Exception:
+                                media_path = None
+
+                            if media_path is not None and media_path.exists() and media_path.is_file():
+                                sub_path = _best_subtitle_sidecar(media_path)
+                                if sub_path is not None:
+                                    sub_text = _read_text_file(sub_path)
+                                    if sub_text:
+                                        notes = po.get("notes")
+                                        if not isinstance(notes, dict):
+                                            notes = {}
+                                        notes["sub"] = sub_text
+                                        po["notes"] = notes
+
+                        pipe_objects.append(po)
 
                     # If this is a clip download, decorate titles/tags so the title: tag is clip-based.
                     # Relationship tags are only added when multiple clips exist.
@@ -1868,6 +2184,95 @@ class Download_Media(Cmdlet):
                     debug("✓ Downloaded and emitted")
 
                 except DownloadError as e:
+                    # Special-case yt-dlp format errors: show a selectable format list table so
+                    # the user can pick a working format_id and continue the pipeline via @N.
+                    cause = getattr(e, "__cause__", None)
+                    detail = ""
+                    try:
+                        detail = str(cause or "")
+                    except Exception:
+                        detail = ""
+
+                    if "requested format is not available" in (detail or "").lower() and mode != "audio":
+                        formats = list_formats(url, no_playlist=False, playlist_items=actual_playlist_items)
+                        if formats:
+                            formats_to_show = formats
+
+                            table = ResultTable()
+                            table.title = f"Available formats for {url}"
+                            table.set_source_command("download-media", [str(a) for a in (args or [])])
+
+                            results_list: List[Dict[str, Any]] = []
+                            for idx, fmt in enumerate(formats_to_show, 1):
+                                resolution = fmt.get("resolution", "")
+                                ext = fmt.get("ext", "")
+                                vcodec = fmt.get("vcodec", "none")
+                                acodec = fmt.get("acodec", "none")
+                                filesize = fmt.get("filesize")
+                                format_id = fmt.get("format_id", "")
+
+                                selection_format_id = format_id
+                                try:
+                                    if vcodec != "none" and acodec == "none" and format_id:
+                                        selection_format_id = f"{format_id}+ba"
+                                except Exception:
+                                    selection_format_id = format_id
+
+                                size_str = ""
+                                if filesize:
+                                    try:
+                                        size_mb = float(filesize) / (1024 * 1024)
+                                        size_str = f"{size_mb:.1f}MB"
+                                    except Exception:
+                                        size_str = ""
+
+                                desc_parts: List[str] = []
+                                if resolution and resolution != "audio only":
+                                    desc_parts.append(str(resolution))
+                                if ext:
+                                    desc_parts.append(str(ext).upper())
+                                if vcodec != "none":
+                                    desc_parts.append(f"v:{vcodec}")
+                                if acodec != "none":
+                                    desc_parts.append(f"a:{acodec}")
+                                if size_str:
+                                    desc_parts.append(size_str)
+                                format_desc = " | ".join(desc_parts)
+
+                                format_dict: Dict[str, Any] = {
+                                    "table": "download-media",
+                                    "title": f"Format {format_id}",
+                                    "url": url,
+                                    "target": url,
+                                    "detail": format_desc,
+                                    "media_kind": "format",
+                                    "columns": [
+                                        ("#", str(idx)),
+                                        ("ID", format_id),
+                                        ("Resolution", resolution or "N/A"),
+                                        ("Ext", ext),
+                                        ("Video", vcodec),
+                                        ("Audio", acodec),
+                                        ("Size", size_str or "N/A"),
+                                    ],
+                                    "full_metadata": {
+                                        "format_id": format_id,
+                                        "url": url,
+                                        "item_selector": selection_format_id,
+                                    },
+                                    "_selection_args": ["-format", selection_format_id],
+                                }
+
+                                results_list.append(format_dict)
+                                table.add_result(format_dict)
+
+                            pipeline_context.set_current_stage_table(table)
+                            pipeline_context.set_last_result_table(table, results_list)
+
+                            # Returning 0 with no emits lets the CLI pause the pipeline for @N selection.
+                            log("Requested format is not available; select a working format with @N", file=sys.stderr)
+                            return 0
+
                     log(f"Download failed for {url}: {e}", file=sys.stderr)
                 except Exception as e:
                     log(f"Error processing {url}: {e}", file=sys.stderr)

cmdlet/download_torrent.py

@@ -15,9 +15,9 @@ from pathlib import Path
 from typing import Any, Dict, Optional, Sequence
 
 from SYS.logger import log
-from ._shared import Cmdlet, CmdletArg, parse_cmdlet_args
+from . import _shared as sh
 
-class Download_Torrent(Cmdlet):
+class Download_Torrent(sh.Cmdlet):
     """Class-based download-torrent cmdlet with self-registration."""
 
     def __init__(self) -> None:
@@ -27,10 +27,10 @@ class Download_Torrent(Cmdlet):
             usage="download-torrent <magnet|.torrent> [options]",
             alias=["torrent", "magnet"],
             arg=[
-                CmdletArg(name="magnet", type="string", required=False, description="Magnet link or .torrent file/URL", variadic=True),
-                CmdletArg(name="output", type="string", description="Output directory for downloaded files"),
-                CmdletArg(name="wait", type="float", description="Wait time (seconds) for magnet processing timeout"),
-                CmdletArg(name="background", type="flag", alias="bg", description="Start download in background"),
+                sh.CmdletArg(name="magnet", type="string", required=False, description="Magnet link or .torrent file/URL", variadic=True),
+                sh.CmdletArg(name="output", type="string", description="Output directory for downloaded files"),
+                sh.CmdletArg(name="wait", type="float", description="Wait time (seconds) for magnet processing timeout"),
+                sh.CmdletArg(name="background", type="flag", alias="bg", description="Start download in background"),
             ],
             detail=["Download torrents/magnets via AllDebrid API."],
             exec=self.run,
@@ -38,7 +38,7 @@ class Download_Torrent(Cmdlet):
         self.register()
 
     def run(self, result: Any, args: Sequence[str], config: Dict[str, Any]) -> int:
-        parsed = parse_cmdlet_args(args, self)
+        parsed = sh.parse_cmdlet_args(args, self)
         magnet_args = parsed.get("magnet", [])
         output_dir = Path(parsed.get("output") or Path.home() / "Downloads")
         wait_timeout = int(float(parsed.get("wait", 600)))

cmdlet/get_file.py

@@ -9,13 +9,13 @@ import subprocess
 import webbrowser
 
 import pipeline as ctx
-from ._shared import Cmdlet, CmdletArg, SharedArgs, parse_cmdlet_args, get_field, normalize_hash
+from . import _shared as sh
 from SYS.logger import log, debug
 from Store import Store
 from config import resolve_output_dir
 
 
-class Get_File(Cmdlet):
+class Get_File(sh.Cmdlet):
     """Export files to local path via hash+store."""
     
     def __init__(self) -> None:
@@ -25,10 +25,10 @@ class Get_File(Cmdlet):
             summary="Export file to local path",
             usage="@1 | get-file -path C:\\Downloads",
             arg=[
-                SharedArgs.HASH,
-                SharedArgs.STORE,
-                SharedArgs.PATH,
-                CmdletArg("name", description="Output filename (default: from metadata title)"),
+                sh.SharedArgs.HASH,
+                sh.SharedArgs.STORE,
+                sh.SharedArgs.PATH,
+                sh.CmdletArg("name", description="Output filename (default: from metadata title)"),
             ],
             detail=[
                 "- Exports file from storage backend to local path",
@@ -42,12 +42,12 @@ class Get_File(Cmdlet):
     def run(self, result: Any, args: Sequence[str], config: Dict[str, Any]) -> int:
         """Export file via hash+store backend."""
         debug(f"[get-file] run() called with result type: {type(result)}")
-        parsed = parse_cmdlet_args(args, self)
+        parsed = sh.parse_cmdlet_args(args, self)
         debug(f"[get-file] parsed args: {parsed}")
         
         # Extract hash and store from result or args
-        file_hash = parsed.get("hash") or get_field(result, "hash")
-        store_name = parsed.get("store") or get_field(result, "store")
+        file_hash = parsed.get("hash") or sh.get_field(result, "hash")
+        store_name = parsed.get("store") or sh.get_field(result, "store")
         output_path = parsed.get("path")
         output_name = parsed.get("name")
         
@@ -62,7 +62,7 @@ class Get_File(Cmdlet):
             return 1
         
         # Normalize hash
-        file_hash = normalize_hash(file_hash)
+        file_hash = sh.normalize_hash(file_hash)
         if not file_hash:
             log("Error: Invalid hash format")
             return 1
@@ -84,9 +84,9 @@ class Get_File(Cmdlet):
 
         def resolve_display_title() -> str:
             candidates = [
-                get_field(result, "title"),
-                get_field(result, "name"),
-                get_field(result, "filename"),
+                sh.get_field(result, "title"),
+                sh.get_field(result, "name"),
+                sh.get_field(result, "filename"),
                 (metadata.get("title") if isinstance(metadata, dict) else None),
                 (metadata.get("name") if isinstance(metadata, dict) else None),
                 (metadata.get("filename") if isinstance(metadata, dict) else None),

cmdlet/get_metadata.py

@@ -7,7 +7,13 @@ import sys
 from SYS.logger import log
 from pathlib import Path
 
-from ._shared import Cmdlet, CmdletArg, SharedArgs, parse_cmdlet_args, get_field
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+parse_cmdlet_args = sh.parse_cmdlet_args
+get_field = sh.get_field
 import pipeline as ctx
 from result_table import ResultTable
 
@@ -74,9 +80,15 @@ class Get_Metadata(Cmdlet):
                          hash_value: Optional[str], pages: Optional[int] = None) -> Dict[str, Any]:
         """Build a table row dict with metadata fields."""
         size_mb = None
-        if isinstance(size_bytes, int):
+        size_int: Optional[int] = None
+        if size_bytes is not None:
             try:
-                size_mb = int(size_bytes / (1024 * 1024))
+                size_int = int(size_bytes)
+            except Exception:
+                size_int = None
+        if isinstance(size_int, int):
+            try:
+                size_mb = int(size_int / (1024 * 1024))
             except Exception:
                 size_mb = None
 
@@ -105,7 +117,7 @@ class Get_Metadata(Cmdlet):
             "path": path,
             "store": store,
             "mime": mime,
-            "size_bytes": size_bytes,
+            "size_bytes": size_int,
             "duration_seconds": dur_int,
             "pages": pages_int,
             "imported_ts": imported_ts,
@@ -237,8 +249,8 @@ class Get_Metadata(Cmdlet):
                 pages=pages,
             )
             
-            table_title = title
-            table = ResultTable(table_title).init_command("get-metadata", list(args))
+            table_title = f"get-metadata: {title}" if title else "get-metadata"
+            table = ResultTable(table_title).init_command(table_title, "get-metadata", list(args))
             self._add_table_body_row(table, row)
             ctx.set_last_result_table_overlay(table, [row], row)
             ctx.emit(row)

cmdlet/get_note.py

@@ -7,15 +7,15 @@ import sys
 from SYS.logger import log
 
 import pipeline as ctx
-from ._shared import (
-    Cmdlet,
-    CmdletArg,
-    SharedArgs,
-    normalize_hash,
-    parse_cmdlet_args,
-    normalize_result_input,
-    should_show_help,
-)
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+normalize_hash = sh.normalize_hash
+parse_cmdlet_args = sh.parse_cmdlet_args
+normalize_result_input = sh.normalize_result_input
+should_show_help = sh.should_show_help
 from Store import Store
 from SYS.utils import sha256_file
 

cmdlet/get_relationship.py

@@ -10,7 +10,17 @@ from SYS.logger import log
 import models
 import pipeline as ctx
 from API import HydrusNetwork as hydrus_wrapper
-from ._shared import Cmdlet, CmdletArg, SharedArgs, normalize_hash, fmt_bytes, get_hash_for_operation, fetch_hydrus_metadata, should_show_help, get_field
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+normalize_hash = sh.normalize_hash
+fmt_bytes = sh.fmt_bytes
+get_hash_for_operation = sh.get_hash_for_operation
+fetch_hydrus_metadata = sh.fetch_hydrus_metadata
+should_show_help = sh.should_show_help
+get_field = sh.get_field
 from API.folder import API_folder_store
 from config import get_local_storage_path
 from result_table import ResultTable
@@ -224,13 +234,14 @@ def _run(result: Any, _args: Sequence[str], config: Dict[str, Any]) -> int:
         try:
             client = None
             store_label = "hydrus"
+            backend_obj = None
             if store_name:
                 # Store specified: do not fall back to a global/default Hydrus client.
                 store_label = str(store_name)
                 try:
                     store = Store(config)
-                    backend = store[str(store_name)]
-                    candidate = getattr(backend, "_client", None)
+                    backend_obj = store[str(store_name)]
+                    candidate = getattr(backend_obj, "_client", None)
                     if candidate is not None and hasattr(candidate, "get_file_relationships"):
                         client = candidate
                 except Exception:
@@ -241,6 +252,74 @@ def _run(result: Any, _args: Sequence[str], config: Dict[str, Any]) -> int:
             else:
                 client = hydrus_wrapper.get_client(config)
 
+            def _resolve_related_title(rel_hash: str) -> str:
+                """Best-effort resolve a Hydrus hash to a human title.
+
+                Preference order:
+                - title: tag from the backend (fast path)
+                - Hydrus metadata tags via fetch_hydrus_metadata
+                - fallback to short hash
+                """
+                h = normalize_hash(rel_hash)
+                if not h:
+                    return str(rel_hash)
+
+                # Prefer backend tag extraction when available.
+                if backend_obj is not None and hasattr(backend_obj, "get_tag"):
+                    try:
+                        tag_result = backend_obj.get_tag(h)
+                        tags = tag_result[0] if isinstance(tag_result, tuple) and tag_result else tag_result
+                        if isinstance(tags, list):
+                            for t in tags:
+                                if isinstance(t, str) and t.lower().startswith("title:"):
+                                    val = t.split(":", 1)[1].strip()
+                                    if val:
+                                        return val
+                    except Exception:
+                        pass
+
+                # Fallback: fetch minimal metadata and scan for a title tag.
+                try:
+                    meta, _ = fetch_hydrus_metadata(
+                        config,
+                        h,
+                        store_name=store_label if store_name else None,
+                        hydrus_client=client,
+                        include_service_keys_to_tags=True,
+                        include_file_url=False,
+                        include_duration=False,
+                        include_size=False,
+                        include_mime=False,
+                    )
+                    if isinstance(meta, dict):
+                        tags_payload = meta.get("tags")
+                        tag_candidates: list[str] = []
+                        if isinstance(tags_payload, dict):
+                            for svc_data in tags_payload.values():
+                                if not isinstance(svc_data, dict):
+                                    continue
+                                storage = svc_data.get("storage_tags")
+                                if isinstance(storage, dict):
+                                    for group in storage.values():
+                                        if isinstance(group, list):
+                                            tag_candidates.extend([str(x) for x in group if isinstance(x, str)])
+                                display = svc_data.get("display_tags")
+                                if isinstance(display, list):
+                                    tag_candidates.extend([str(x) for x in display if isinstance(x, str)])
+                        flat = meta.get("tags_flat")
+                        if isinstance(flat, list):
+                            tag_candidates.extend([str(x) for x in flat if isinstance(x, str)])
+
+                        for t in tag_candidates:
+                            if isinstance(t, str) and t.lower().startswith("title:"):
+                                val = t.split(":", 1)[1].strip()
+                                if val:
+                                    return val
+                except Exception:
+                    pass
+
+                return h[:16] + "..."
+
             if client:
                 rel = client.get_file_relationships(hash_hex)
                 if rel:
@@ -274,7 +353,7 @@ def _run(result: Any, _args: Sequence[str], config: Dict[str, Any]) -> int:
                                         found_relationships.append({
                                             "hash": king_hash,
                                             "type": "king",
-                                            "title": king_hash,
+                                            "title": _resolve_related_title(king_hash),
                                             "path": None,
                                             "store": store_label,
                                         })
@@ -292,7 +371,7 @@ def _run(result: Any, _args: Sequence[str], config: Dict[str, Any]) -> int:
                                         found_relationships.append({
                                             "hash": rel_hash_norm,
                                             "type": rel_name,
-                                            "title": rel_hash_norm,  # Can't resolve title easily without another API call
+                                            "title": _resolve_related_title(rel_hash_norm),
                                             "path": None,
                                             "store": store_label,
                                         })
@@ -304,7 +383,7 @@ def _run(result: Any, _args: Sequence[str], config: Dict[str, Any]) -> int:
                                         found_relationships.append({
                                             "hash": rel_hash_norm,
                                             "type": rel_name,
-                                            "title": rel_hash_norm,
+                                            "title": _resolve_related_title(rel_hash_norm),
                                             "path": None,
                                             "store": store_label,
                                         })

cmdlet/get_tag.py

@@ -27,7 +27,15 @@ from typing import Any, Dict, List, Optional, Sequence, Tuple
 import pipeline as ctx
 from API import HydrusNetwork
 from API.folder import read_sidecar, write_sidecar, find_sidecar, API_folder_store
-from ._shared import normalize_hash, looks_like_hash, Cmdlet, CmdletArg, SharedArgs, parse_cmdlet_args, get_field
+from . import _shared as sh
+
+normalize_hash = sh.normalize_hash
+looks_like_hash = sh.looks_like_hash
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+parse_cmdlet_args = sh.parse_cmdlet_args
+get_field = sh.get_field
 from config import get_local_storage_path
 
 

cmdlet/get_url.py

@@ -5,7 +5,15 @@ from typing import Any, Dict, List, Sequence
 import sys
 
 import pipeline as ctx
-from ._shared import Cmdlet, SharedArgs, parse_cmdlet_args, get_field, normalize_hash
+from . import _shared as sh
+
+Cmdlet, SharedArgs, parse_cmdlet_args, get_field, normalize_hash = (
+    sh.Cmdlet,
+    sh.SharedArgs,
+    sh.parse_cmdlet_args,
+    sh.get_field,
+    sh.normalize_hash,
+)
 from SYS.logger import log
 from Store import Store
 

cmdlet/merge_file.py

@@ -12,17 +12,17 @@ import re as _re
 
 from config import resolve_output_dir
 
-from ._shared import (
-    Cmdlet,
-    CmdletArg,
-    create_pipe_object_result,
-    get_field,
-    get_pipe_object_hash,
-    get_pipe_object_path,
-    normalize_result_input,
-    parse_cmdlet_args,
-    should_show_help,
-)
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+create_pipe_object_result = sh.create_pipe_object_result
+get_field = sh.get_field
+get_pipe_object_hash = sh.get_pipe_object_hash
+get_pipe_object_path = sh.get_pipe_object_path
+normalize_result_input = sh.normalize_result_input
+parse_cmdlet_args = sh.parse_cmdlet_args
+should_show_help = sh.should_show_help
 
 import pipeline as ctx
 

cmdlet/screen_shot.py

@@ -20,7 +20,16 @@ from urllib.parse import urlsplit, quote, urljoin
 from SYS.logger import log, debug
 from API.HTTP import HTTPClient
 from SYS.utils import ensure_directory, unique_path, unique_preserve_order
-from ._shared import Cmdlet, CmdletArg, SharedArgs, create_pipe_object_result, normalize_result_input, should_show_help, get_field
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+SharedArgs = sh.SharedArgs
+create_pipe_object_result = sh.create_pipe_object_result
+normalize_result_input = sh.normalize_result_input
+should_show_help = sh.should_show_help
+get_field = sh.get_field
+parse_cmdlet_args = sh.parse_cmdlet_args
 import pipeline as pipeline_context
 
 # ============================================================================
@@ -33,20 +42,7 @@ import pipeline as pipeline_context
 # Playwright & Screenshot Dependencies
 # ============================================================================
 
-try:
-    from playwright.sync_api import (
-        TimeoutError as PlaywrightTimeoutError,
-        sync_playwright,
-    )
-    HAS_PLAYWRIGHT = True
-except Exception:
-    HAS_PLAYWRIGHT = False
-    PlaywrightTimeoutError = TimeoutError  # type: ignore
-
-    def sync_playwright(*_args: Any, **_kwargs: Any) -> Any:  # type: ignore
-        raise RuntimeError(
-            "playwright is required for screenshot capture; install with: pip install playwright; then: playwright install"
-        )
+from tool.playwright import HAS_PLAYWRIGHT, PlaywrightTimeoutError, PlaywrightTool
 
 try:
     from config import resolve_output_dir
@@ -128,6 +124,7 @@ class ScreenshotOptions:
     prefer_platform_target: bool = False
     target_selectors: Optional[Sequence[str]] = None
     selector_timeout_ms: int = 10_000
+    playwright_tool: Optional[PlaywrightTool] = None
 
 
 @dataclass(slots=True)
@@ -324,142 +321,119 @@ def _prepare_output_path(options: ScreenshotOptions) -> Path:
 def _capture(options: ScreenshotOptions, destination: Path, warnings: List[str]) -> None:
     """Capture screenshot using Playwright."""
     debug(f"[_capture] Starting capture for {options.url} -> {destination}")
-    playwright = None
-    browser = None
-    context = None
     try:
-        debug("Starting Playwright...", flush=True)
-        playwright = sync_playwright().start()
-        log("Launching Chromium browser...", flush=True)
+        tool = options.playwright_tool or PlaywrightTool({})
+        tool.debug_dump()
+
+        log("Launching browser...", flush=True)
         format_name = _normalise_format(options.output_format)
         headless = options.headless or format_name == "pdf"
         debug(f"[_capture] Format: {format_name}, Headless: {headless}")
         
         if format_name == "pdf" and not options.headless:
             warnings.append("pdf output requires headless Chromium; overriding headless mode")
-        browser = playwright.chromium.launch(
-            headless=headless,
-            args=["--disable-blink-features=AutomationControlled"],
-        )
-        log("Creating browser context...", flush=True)
-        context = browser.new_context(
-            user_agent=USER_AGENT,
-            viewport=DEFAULT_VIEWPORT,
-            ignore_https_errors=True,
-        )
-        page = context.new_page()
-        log(f"Navigating to {options.url}...", flush=True)
-        try:
-            page.goto(options.url, timeout=90_000, wait_until="domcontentloaded")
-            log("Page loaded successfully", flush=True)
-        except PlaywrightTimeoutError:
-            warnings.append("navigation timeout; capturing current page state")
-            log("Navigation timeout; proceeding with current state", flush=True)
-        
-        # Skip article lookup by default (wait_for_article defaults to False)
-        if options.wait_for_article:
+
+        with tool.open_page(headless=headless) as page:
+            log(f"Navigating to {options.url}...", flush=True)
             try:
-                log("Waiting for article element...", flush=True)
-                page.wait_for_selector("article", timeout=10_000)
-                log("Article element found", flush=True)
+                tool.goto(page, options.url)
+                log("Page loaded successfully", flush=True)
             except PlaywrightTimeoutError:
-                warnings.append("<article> selector not found; capturing fallback")
-                log("Article element not found; using fallback", flush=True)
-        
-        if options.wait_after_load > 0:
-            log(f"Waiting {options.wait_after_load}s for page stabilization...", flush=True)
-            time.sleep(min(10.0, max(0.0, options.wait_after_load)))
-        if options.replace_video_posters:
-            log("Replacing video elements with posters...", flush=True)
-            page.evaluate(
-                """
-                    document.querySelectorAll('video').forEach(v => {
-                        if (v.poster) {
-                            const img = document.createElement('img');
-                            img.src = v.poster;
-                            img.style.maxWidth = '100%';
-                            img.style.borderRadius = '12px';
-                            v.replaceWith(img);
-                        }
-                    });
-                """
-            )
-        # Attempt platform-specific target capture if requested (and not PDF)
-        element_captured = False
-        if options.prefer_platform_target and format_name != "pdf":
-            log("Attempting platform-specific content capture...", flush=True)
-            try:
-                _platform_preprocess(options.url, page, warnings)
-            except Exception as e:
-                debug(f"[_capture] Platform preprocess failed: {e}")
-                pass
-            selectors = list(options.target_selectors or [])
-            if not selectors:
-                selectors = _selectors_for_url(options.url)
+                warnings.append("navigation timeout; capturing current page state")
+                log("Navigation timeout; proceeding with current state", flush=True)
             
-            debug(f"[_capture] Trying selectors: {selectors}")
-            for sel in selectors:
+            # Skip article lookup by default (wait_for_article defaults to False)
+            if options.wait_for_article:
                 try:
-                    log(f"Trying selector: {sel}", flush=True)
-                    el = page.wait_for_selector(sel, timeout=max(0, int(options.selector_timeout_ms)))
+                    log("Waiting for article element...", flush=True)
+                    page.wait_for_selector("article", timeout=10_000)
+                    log("Article element found", flush=True)
                 except PlaywrightTimeoutError:
-                    log(f"Selector not found: {sel}", flush=True)
-                    continue
+                    warnings.append("<article> selector not found; capturing fallback")
+                    log("Article element not found; using fallback", flush=True)
+            
+            if options.wait_after_load > 0:
+                log(f"Waiting {options.wait_after_load}s for page stabilization...", flush=True)
+                time.sleep(min(10.0, max(0.0, options.wait_after_load)))
+            if options.replace_video_posters:
+                log("Replacing video elements with posters...", flush=True)
+                page.evaluate(
+                    """
+                        document.querySelectorAll('video').forEach(v => {
+                            if (v.poster) {
+                                const img = document.createElement('img');
+                                img.src = v.poster;
+                                img.style.maxWidth = '100%';
+                                img.style.borderRadius = '12px';
+                                v.replaceWith(img);
+                            }
+                        });
+                    """
+                )
+            # Attempt platform-specific target capture if requested (and not PDF)
+            element_captured = False
+            if options.prefer_platform_target and format_name != "pdf":
+                log("Attempting platform-specific content capture...", flush=True)
                 try:
-                    if el is not None:
-                        log(f"Found element with selector: {sel}", flush=True)
-                        try:
-                            el.scroll_into_view_if_needed(timeout=1000)
-                        except Exception:
-                            pass
-                        log(f"Capturing element to {destination}...", flush=True)
-                        el.screenshot(path=str(destination), type=("jpeg" if format_name == "jpeg" else None))
-                        element_captured = True
-                        log("Element captured successfully", flush=True)
-                        break
-                except Exception as exc:
-                    warnings.append(f"element capture failed for '{sel}': {exc}")
-                    log(f"Failed to capture element: {exc}", flush=True)
-        # Fallback to default capture paths
-        if element_captured:
-            pass
-        elif format_name == "pdf":
-            log("Generating PDF...", flush=True)
-            page.emulate_media(media="print")
-            page.pdf(path=str(destination), print_background=True)
-            log(f"PDF saved to {destination}", flush=True)
-        else:
-            log(f"Capturing full page to {destination}...", flush=True)
-            screenshot_kwargs: Dict[str, Any] = {"path": str(destination)}
-            if format_name == "jpeg":
-                screenshot_kwargs["type"] = "jpeg"
-                screenshot_kwargs["quality"] = 90
-            if options.full_page:
-                page.screenshot(full_page=True, **screenshot_kwargs)
+                    _platform_preprocess(options.url, page, warnings)
+                except Exception as e:
+                    debug(f"[_capture] Platform preprocess failed: {e}")
+                    pass
+                selectors = list(options.target_selectors or [])
+                if not selectors:
+                    selectors = _selectors_for_url(options.url)
+                
+                debug(f"[_capture] Trying selectors: {selectors}")
+                for sel in selectors:
+                    try:
+                        log(f"Trying selector: {sel}", flush=True)
+                        el = page.wait_for_selector(sel, timeout=max(0, int(options.selector_timeout_ms)))
+                    except PlaywrightTimeoutError:
+                        log(f"Selector not found: {sel}", flush=True)
+                        continue
+                    try:
+                        if el is not None:
+                            log(f"Found element with selector: {sel}", flush=True)
+                            try:
+                                el.scroll_into_view_if_needed(timeout=1000)
+                            except Exception:
+                                pass
+                            log(f"Capturing element to {destination}...", flush=True)
+                            el.screenshot(path=str(destination), type=("jpeg" if format_name == "jpeg" else None))
+                            element_captured = True
+                            log("Element captured successfully", flush=True)
+                            break
+                    except Exception as exc:
+                        warnings.append(f"element capture failed for '{sel}': {exc}")
+                        log(f"Failed to capture element: {exc}", flush=True)
+            # Fallback to default capture paths
+            if element_captured:
+                pass
+            elif format_name == "pdf":
+                log("Generating PDF...", flush=True)
+                page.emulate_media(media="print")
+                page.pdf(path=str(destination), print_background=True)
+                log(f"PDF saved to {destination}", flush=True)
             else:
-                article = page.query_selector("article")
-                if article is not None:
-                    article_kwargs = dict(screenshot_kwargs)
-                    article_kwargs.pop("full_page", None)
-                    article.screenshot(**article_kwargs)
+                log(f"Capturing full page to {destination}...", flush=True)
+                screenshot_kwargs: Dict[str, Any] = {"path": str(destination)}
+                if format_name == "jpeg":
+                    screenshot_kwargs["type"] = "jpeg"
+                    screenshot_kwargs["quality"] = 90
+                if options.full_page:
+                    page.screenshot(full_page=True, **screenshot_kwargs)
                 else:
-                    page.screenshot(**screenshot_kwargs)
-            log(f"Screenshot saved to {destination}", flush=True)
+                    article = page.query_selector("article")
+                    if article is not None:
+                        article_kwargs = dict(screenshot_kwargs)
+                        article_kwargs.pop("full_page", None)
+                        article.screenshot(**article_kwargs)
+                    else:
+                        page.screenshot(**screenshot_kwargs)
+                log(f"Screenshot saved to {destination}", flush=True)
     except Exception as exc:
         debug(f"[_capture] Exception: {exc}")
         raise ScreenshotError(f"Failed to capture screenshot: {exc}") from exc
-    finally:
-        log("Cleaning up browser resources...", flush=True)
-        with contextlib.suppress(Exception):
-            if context is not None:
-                context.close()
-        with contextlib.suppress(Exception):
-            if browser is not None:
-                browser.close()
-        with contextlib.suppress(Exception):
-            if playwright is not None:
-                playwright.stop()
-        log("Cleanup complete", flush=True)
 
 
 def _capture_screenshot(options: ScreenshotOptions) -> ScreenshotResult:
@@ -511,8 +485,6 @@ def _run(result: Any, args: Sequence[str], config: Dict[str, Any]) -> int:
     Screenshots are created using Playwright and marked as temporary
     so they can be cleaned up later with the cleanup cmdlet.
     """
-    from ._shared import parse_cmdlet_args
-    
     debug(f"[_run] screen-shot invoked with args: {args}")
 
     # Help check
@@ -534,6 +506,19 @@ def _run(result: Any, args: Sequence[str], config: Dict[str, Any]) -> int:
     parsed = parse_cmdlet_args(args, CMDLET)
     
     format_value = parsed.get("format")
+    if not format_value:
+        # Default format can be set via config.conf tool block:
+        #   [tool=playwright]
+        #   format="pdf"
+        try:
+            tool_cfg = config.get("tool", {}) if isinstance(config, dict) else {}
+            pw_cfg = tool_cfg.get("playwright") if isinstance(tool_cfg, dict) else None
+            if isinstance(pw_cfg, dict):
+                format_value = pw_cfg.get("format")
+        except Exception:
+            pass
+    if not format_value:
+        format_value = "png"
     storage_value = parsed.get("storage")
     selector_arg = parsed.get("selector")
     selectors = [selector_arg] if selector_arg else []
@@ -669,6 +654,7 @@ def _run(result: Any, args: Sequence[str], config: Dict[str, Any]) -> int:
                 prefer_platform_target=False,
                 wait_for_article=False,
                 full_page=True,
+                playwright_tool=PlaywrightTool(config),
             )
             
             screenshot_result = _capture_screenshot(options)

cmdlet/search_provider.py

@@ -10,7 +10,13 @@ import importlib
 from SYS.logger import log, debug
 from ProviderCore.registry import get_search_provider, list_search_providers
 
-from ._shared import Cmdlet, CmdletArg, should_show_help
+from . import _shared as sh
+
+Cmdlet, CmdletArg, should_show_help = (
+    sh.Cmdlet,
+    sh.CmdletArg,
+    sh.should_show_help,
+)
 import pipeline as ctx
 
 # Optional dependencies

cmdlet/search_store.py

@@ -10,7 +10,17 @@ import sys
 
 from SYS.logger import log, debug
 
-from ._shared import Cmdlet, CmdletArg, SharedArgs, get_field, should_show_help, normalize_hash, first_title_tag
+from . import _shared as sh
+
+Cmdlet, CmdletArg, SharedArgs, get_field, should_show_help, normalize_hash, first_title_tag = (
+    sh.Cmdlet,
+    sh.CmdletArg,
+    sh.SharedArgs,
+    sh.get_field,
+    sh.should_show_help,
+    sh.normalize_hash,
+    sh.first_title_tag,
+)
 import pipeline as ctx
 
 
@@ -209,6 +219,10 @@ class Search_Store(Cmdlet):
                     table_title += f" [{storage_backend}]"
 
                 table = ResultTable(table_title)
+                try:
+                    table.set_source_command("search-store", list(args_list))
+                except Exception:
+                    pass
                 if hash_query:
                     try:
                         table.set_preserve_order(True)
@@ -309,6 +323,11 @@ class Search_Store(Cmdlet):
                                 ext_val = Path(path_str).suffix
                             except Exception:
                                 ext_val = None
+                        if not ext_val and title:
+                            try:
+                                ext_val = Path(str(title)).suffix
+                            except Exception:
+                                ext_val = None
 
                         size_bytes = meta_obj.get("size")
                         if size_bytes is None:
@@ -333,6 +352,20 @@ class Search_Store(Cmdlet):
                         ctx.emit(payload)
 
                     if found_any:
+                        # Title should reflect the command, query, and only stores present in the table.
+                        store_counts: "OrderedDict[str, int]" = OrderedDict()
+                        for row_item in results_list:
+                            store_val = str(row_item.get("store") or "").strip()
+                            if not store_val:
+                                continue
+                            if store_val not in store_counts:
+                                store_counts[store_val] = 0
+                            store_counts[store_val] += 1
+
+                        counts_part = " ".join(f"{name}:{count}" for name, count in store_counts.items() if count > 0)
+                        base_title = f"search-store: {query}".strip()
+                        table.title = f"{base_title} | {counts_part}" if counts_part else base_title
+
                         ctx.set_last_result_table(table, results_list)
                         db.append_worker_stdout(worker_id, json.dumps(results_list, indent=2))
                         db.update_worker_status(worker_id, 'completed')
@@ -377,28 +410,6 @@ class Search_Store(Cmdlet):
                             log(f"Backend {backend_name} search failed: {exc}", file=sys.stderr)
                     results = all_results[:limit]
 
-                def _format_storage_label(name: str) -> str:
-                    clean = str(name or "").strip()
-                    if not clean:
-                        return "Unknown"
-                    return clean.replace("_", " ").title()
-
-                storage_counts: OrderedDict[str, int] = OrderedDict((name, 0) for name in searched_backends)
-                for item in results or []:
-                    store = get_field(item, "store")
-                    if not store:
-                        continue
-                    key = str(store).lower()
-                    if key not in storage_counts:
-                        storage_counts[key] = 0
-                    storage_counts[key] += 1
-
-                if storage_counts or query:
-                    display_counts = OrderedDict((_format_storage_label(name), count) for name, count in storage_counts.items())
-                    summary_line = table.set_storage_summary(display_counts, query, inline=True)
-                    if summary_line:
-                        table.title = summary_line
-
                 if results:
                     for item in results:
                         def _as_dict(obj: Any) -> Dict[str, Any]:
@@ -428,6 +439,20 @@ class Search_Store(Cmdlet):
                         results_list.append(normalized)
                         ctx.emit(normalized)
 
+                    # Title should reflect the command, query, and only stores present in the table.
+                    store_counts: "OrderedDict[str, int]" = OrderedDict()
+                    for row_item in results_list:
+                        store_val = str(row_item.get("store") or "").strip()
+                        if not store_val:
+                            continue
+                        if store_val not in store_counts:
+                            store_counts[store_val] = 0
+                        store_counts[store_val] += 1
+
+                    counts_part = " ".join(f"{name}:{count}" for name, count in store_counts.items() if count > 0)
+                    base_title = f"search-store: {query}".strip()
+                    table.title = f"{base_title} | {counts_part}" if counts_part else base_title
+
                     ctx.set_last_result_table(table, results_list)
                     db.append_worker_stdout(worker_id, json.dumps(results_list, indent=2))
                 else:

cmdlet/trim_file.py

@@ -11,14 +11,14 @@ import re
 
 from SYS.logger import log, debug
 from SYS.utils import sha256_file
-from ._shared import (
-    Cmdlet,
-    CmdletArg,
-    parse_cmdlet_args,
-    normalize_result_input,
-    extract_tag_from_result,
-    extract_title_from_result
-)
+from . import _shared as sh
+
+Cmdlet = sh.Cmdlet
+CmdletArg = sh.CmdletArg
+parse_cmdlet_args = sh.parse_cmdlet_args
+normalize_result_input = sh.normalize_result_input
+extract_tag_from_result = sh.extract_tag_from_result
+extract_title_from_result = sh.extract_title_from_result
 import pipeline as ctx
 
 CMDLET = Cmdlet(