diff --git a/pyproject.toml b/pyproject.toml
index 884b452c..19d853f5 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -44,6 +44,7 @@ dependencies = [
     "fastmcp>=2.14.1",
     "httpx>=0.28.1",
     "jieba>=0.42.1",
+    "rjieba>=0.1.11",
     "loguru>=0.7.3",
     "mcp>=1.25.0",
     "networkx>=3.4",
diff --git a/reme4/application.py b/reme4/application.py
index dc2a3fd4..969fcd20 100644
--- a/reme4/application.py
+++ b/reme4/application.py
@@ -20,10 +20,14 @@ class Application(BaseComponent):
 
         vault_path = Path(self.config.vault_dir).absolute()
         vault_path.mkdir(parents=True, exist_ok=True)
-        (vault_path / self.config.metadata_dir).mkdir(parents=True, exist_ok=True)
-        (vault_path / self.config.daily_dir).mkdir(parents=True, exist_ok=True)
-        (vault_path / self.config.digest_dir).mkdir(parents=True, exist_ok=True)
-        (vault_path / self.config.resource_dir).mkdir(parents=True, exist_ok=True)
+        if self.config.metadata_dir:
+            (vault_path / self.config.metadata_dir).mkdir(parents=True, exist_ok=True)
+        if self.config.daily_dir:
+            (vault_path / self.config.daily_dir).mkdir(parents=True, exist_ok=True)
+        if self.config.digest_dir:
+            (vault_path / self.config.digest_dir).mkdir(parents=True, exist_ok=True)
+        if self.config.resource_dir:
+            (vault_path / self.config.resource_dir).mkdir(parents=True, exist_ok=True)
 
         if self.config.enable_logo:
             print_logo(self.config)
@@ -60,15 +64,16 @@ class Application(BaseComponent):
                 self.context.components[component_type][name] = backend_cls(**params)
 
         # Jobs
-        for job_config in self.config.jobs:
+        for name, job_config in self.config.jobs.items():
             if not job_config.backend:
-                raise ValueError(f"Job '{job_config.name}' is missing the required 'backend' field")
+                raise ValueError(f"Job '{name}' is missing the required 'backend' field")
             job_cls = R.get(ComponentEnum.JOB, job_config.backend)
             if not job_cls:
-                raise ValueError(f"Unregistered backend '{job_config.backend}' for job '{job_config.name}'")
+                raise ValueError(f"Unregistered backend '{job_config.backend}' for job '{name}'")
             params = job_config.model_dump()
+            params.setdefault("name", name)
             params["app_context"] = self.context
-            self.context.jobs[job_config.name] = job_cls(**params)
+            self.context.jobs[name] = job_cls(**params)
 
     @property
     def config(self):
diff --git a/reme4/components/application_context.py b/reme4/components/application_context.py
index 90a7f194..0feac294 100644
--- a/reme4/components/application_context.py
+++ b/reme4/components/application_context.py
@@ -1,28 +1,29 @@
 """Application context: shared state container for components, jobs, and service."""
 
+from typing import TYPE_CHECKING
+
 from ..enumeration import ComponentEnum
 from ..schema import ApplicationConfig
 
+if TYPE_CHECKING:
+    from .base_component import BaseComponent
+    from .job import BaseJob
+    from .service import BaseService
+
 
 class ApplicationContext:
-    """Holds the parsed config and instantiated components, jobs, and service.
+    """Passive state container holding parsed config and wired components.
 
-    Acts as a passive state container. The actual wiring (resolving backends from
-    the registry and instantiating each component) is performed by Application.
+    The Application class performs the actual wiring (registry lookups and
+    component instantiation); this class only stores the results so that
+    components, jobs, and the service can find each other at runtime.
     """
 
     def __init__(self, **kwargs):
-        # Parse and validate raw config kwargs into a typed ApplicationConfig.
+        # Parse raw kwargs into a typed, validated config object.
         self.app_config: ApplicationConfig = ApplicationConfig(**kwargs)
 
-        # Local imports to avoid circular dependencies during module init.
-        from .base_component import BaseComponent
-        from .job import BaseJob
-        from .service import BaseService
-
-        # Service endpoint (e.g. HTTP/MCP). Populated by Application.__init__.
-        self.service: BaseService | None = None
-        # Components keyed by type then by user-defined name.
-        self.components: dict[ComponentEnum, dict[str, BaseComponent]] = {}
-        # Jobs keyed by user-defined name.
-        self.jobs: dict[str, BaseJob] = {}
+        # Populated by Application during initialization.
+        self.service: "BaseService | None" = None
+        self.components: dict[ComponentEnum, dict[str, "BaseComponent"]] = {}
+        self.jobs: dict[str, "BaseJob"] = {}
diff --git a/reme4/components/base_component.py b/reme4/components/base_component.py
index ba238cb0..9d8bf50e 100644
--- a/reme4/components/base_component.py
+++ b/reme4/components/base_component.py
@@ -1,4 +1,4 @@
-"""Base class for components."""
+"""Base class for components with async lifecycle and dependency injection."""
 
 import asyncio
 from abc import ABC
@@ -15,7 +15,11 @@ T = TypeVar("T", bound="BaseComponent")
 
 
 class Dependency:
-    """Declared dependency: bind() return value, instance attribute placeholder, and topological-sort edge."""
+    """Placeholder returned by ``BaseComponent.bind`` for an unresolved dependency.
+
+    Resolved into a real component (or None) when the owning component starts.
+    Accessing any attribute before resolution raises a clear error.
+    """
 
     __slots__ = ("ctype", "name", "default_factory", "optional")
 
@@ -36,9 +40,10 @@ class Dependency:
         return f"<unresolved {self.ctype.value}:{self.name}{suffix}>"
 
     def __getattr__(self, item: str) -> Any:
-        # Guard against using the dependency before start() resolves it.
+        # Catches accidental use of the placeholder before start() resolves it.
         raise RuntimeError(
-            f"Dependency {self.ctype.value}:{self.name} accessed before start() (attribute '{item}')",
+            f"Dependency {self.ctype.value}:{self.name} accessed before start() "
+            f"(attribute '{item}')",
         )
 
 
@@ -58,13 +63,14 @@ class BaseComponent(ABC):
         self.backend: str = backend
         self.app_context: "ApplicationContext | None" = app_context
         self.kwargs: dict = dict(kwargs)
-        self.logger = get_logger()
-        if hasattr(self.logger, "bind"):
-            self.logger = self.logger.bind(component=self.name)
+
+        logger = get_logger()
+        self.logger = logger.bind(component=self.name) if hasattr(logger, "bind") else logger
 
         self._is_started: bool = False
         self._lock: asyncio.Lock = asyncio.Lock()
-        # Components created from bind() default_factory in standalone mode (auto-managed lifecycle).
+        # Components created via bind() default_factory in standalone mode;
+        # their lifecycle is owned by this component.
         self._owned: list["BaseComponent"] = []
 
     @property
@@ -82,87 +88,110 @@ class BaseComponent(ABC):
         default_factory: Callable[[], T] | None = None,
         optional: bool = True,
     ) -> T | None:
-        """Declare a dependency on another component; resolved at start(). Empty name → None."""
+        """Declare a dependency on another component.
+
+        Returns a ``Dependency`` placeholder resolved into the real component
+        (or None / a factory-produced instance) when ``start`` runs. An empty
+        `name` short-circuits to None so callers can skip optional wiring.
+        """
         if not name:
             return None
         ctype = getattr(base_cls, "component_type", None)
         if not isinstance(ctype, ComponentEnum) or ctype is ComponentEnum.BASE:
-            raise TypeError(f"{base_cls.__name__} must declare a non-BASE ComponentEnum 'component_type'")
+            raise TypeError(
+                f"{base_cls.__name__} must declare a non-BASE ComponentEnum 'component_type'"
+            )
         return cast(T, Dependency(ctype, name, default_factory, optional))
 
     @property
     def dependencies(self) -> list[Dependency]:
-        """All unresolved bindings declared on this instance."""
+        """All unresolved dependency placeholders on this instance."""
         return [v for v in self.__dict__.values() if isinstance(v, Dependency)]
 
     async def _resolve_bindings(self) -> None:
-        """Replace Dependency placeholders with real components (or default_factory / None for optional)."""
-        for attr, value in list(self.__dict__.items()):
-            if not isinstance(value, Dependency):
-                continue
-            if self.app_context is None:
-                # Standalone mode: factory or (optional → None) or keep placeholder.
-                if value.default_factory is not None:
-                    instance = value.default_factory()
-                    setattr(self, attr, instance)
-                    if isinstance(instance, BaseComponent):
-                        self._owned.append(instance)
-                elif value.optional:
-                    setattr(self, attr, None)
-            else:
-                target = self.app_context.components.get(value.ctype, {}).get(value.name)
-                if target is not None:
-                    setattr(self, attr, target)
-                elif value.optional:
-                    setattr(self, attr, None)
-                else:
-                    raise ValueError(f"{value.ctype.value} '{value.name}' not found.")
+        """Replace every ``Dependency`` attribute with its resolved target."""
+        for attr, dep in list(self.__dict__.items()):
+            if isinstance(dep, Dependency):
+                self._resolve_one(attr, dep)
 
-    # ----- Lookup --------------------------------------------------------
+    def _resolve_one(self, attr: str, dep: Dependency) -> None:
+        """Resolve a single dependency, dispatching by mode."""
+        if self.app_context is None:
+            self._resolve_standalone(attr, dep)
+        else:
+            self._resolve_from_context(attr, dep)
+
+    def _resolve_standalone(self, attr: str, dep: Dependency) -> None:
+        """Standalone mode: use default_factory, or fall back to None when optional.
+
+        Required dependencies without a factory keep the placeholder so any
+        attribute access surfaces a clear error at the call site.
+        """
+        if dep.default_factory is not None:
+            instance = dep.default_factory()
+            setattr(self, attr, instance)
+            if isinstance(instance, BaseComponent):
+                self._owned.append(instance)
+        elif dep.optional:
+            setattr(self, attr, None)
+
+    def _resolve_from_context(self, attr: str, dep: Dependency) -> None:
+        """Context-bound mode: look up the component from ``app_context.components``."""
+        target = self.app_context.components.get(dep.ctype, {}).get(dep.name)
+        if target is not None:
+            setattr(self, attr, target)
+        elif dep.optional:
+            setattr(self, attr, None)
+        else:
+            raise ValueError(f"{dep.ctype.value} '{dep.name}' not found.")
+
+    # ----- Vault path helpers --------------------------------------------
 
     @property
     def vault_path(self) -> Path:
-        """Resolved vault root path from app context or cwd."""
+        """Absolute vault root directory (cwd when no app_context is attached)."""
         if self.app_context is None:
             return Path.cwd()
         return Path(self.app_context.app_config.vault_dir).absolute()
 
     @property
     def vault_metadata_path(self) -> Path:
-        """Resolved metadata directory: vault_path / metadata_dir, or absolute metadata_dir."""
+        """Vault metadata directory: ``<vault>/<metadata_dir>``."""
         if self.app_context is None:
             return Path.cwd() / "metadata"
         return self.vault_path / self.app_context.app_config.metadata_dir
 
     @property
     def component_metadata_path(self) -> Path:
-        """Resolved component metadata directory: vault_metadata_path / component_type."""
+        """Per-component metadata directory under the vault."""
         return self.vault_metadata_path / self.component_type.value
 
     def to_vault_relative(self, path: str | Path) -> str:
-        """Return path relative to vault_path; absolute path string if outside."""
+        """Convert `path` to a vault-relative string; return absolute path when outside."""
         abs_path = Path(path).absolute()
         try:
             return str(abs_path.relative_to(self.vault_path))
         except ValueError:
             return str(abs_path)
 
-    # ----- Lifecycle -----------------------------------------------------
+    # ----- Lifecycle hooks (override in subclasses) ----------------------
 
     async def _start(self) -> None:
-        """Subclass hook: start logic."""
+        """Subclass hook called once after dependencies are resolved."""
 
     async def _close(self) -> None:
-        """Subclass hook: close logic."""
+        """Subclass hook called once during ``close``."""
 
     async def dump(self) -> None:
-        """Persist in-memory state to disk. Override in subclasses that need persistence."""
+        """Persist in-memory state to disk. Override when persistence is needed."""
 
     async def load(self) -> None:
-        """Restore in-memory state from disk. Override in subclasses that need persistence."""
+        """Restore in-memory state from disk. Override when persistence is needed."""
+
+    # ----- Lifecycle control --------------------------------------------
 
     async def start(self) -> None:
-        """Resolve bindings → start owned fallbacks → _start(). No-op if already started."""
+        """Start the component once: resolve deps → start owned → run _start."""
         async with self._lock:
             if self._is_started:
                 return
@@ -173,7 +202,7 @@ class BaseComponent(ABC):
             self._is_started = True
 
     async def close(self) -> None:
-        """_close() → close owned fallbacks in reverse. No-op if not started."""
+        """Close the component once: run _close → close owned in reverse order."""
         async with self._lock:
             if not self._is_started:
                 return
@@ -183,7 +212,7 @@ class BaseComponent(ABC):
             self._is_started = False
 
     async def restart(self) -> None:
-        """Close then start."""
+        """Close then start the component."""
         await self.close()
         await self.start()
 
diff --git a/reme4/components/component_registry.py b/reme4/components/component_registry.py
index c78ceb9c..3e59b88a 100644
--- a/reme4/components/component_registry.py
+++ b/reme4/components/component_registry.py
@@ -1,4 +1,4 @@
-"""Global registry mapping (ComponentEnum, name) -> component class."""
+"""Global registry mapping ``(ComponentEnum, name) -> component class``."""
 
 from typing import Callable, TypeVar, cast
 
@@ -10,7 +10,7 @@ T = TypeVar("T", bound=BaseComponent)
 
 
 class ComponentRegistry:
-    """Two-level registry: component_type -> name -> class.
+    """Two-level registry: ``component_type -> name -> class``.
 
     Supports both direct calls — ``R.register(MyClass, "name")`` — and
     decorator usage — ``@R.register("name")``.
@@ -21,16 +21,20 @@ class ComponentRegistry:
         self.logger = get_logger()
 
     def _do_register(self, cls: type[T], name: str) -> type[T]:
-        """Insert `cls` under its `component_type` group; warn on overwrite."""
+        """Insert `cls` under its ``component_type`` group; warn on overwrite."""
         component_type = getattr(cls, "component_type", None)
         if not isinstance(component_type, ComponentEnum):
-            raise TypeError(f"{cls.__name__} must have a ComponentEnum 'component_type' attribute")
+            raise TypeError(
+                f"{cls.__name__} must have a ComponentEnum 'component_type' attribute"
+            )
         if not name:
             raise ValueError("Component name cannot be empty")
 
         group = self._registry.setdefault(component_type, {})
         if name in group:
-            self.logger.warning(f"Component '{name}' already registered for {component_type}, overwriting")
+            self.logger.warning(
+                f"Component '{name}' already registered for {component_type}, overwriting"
+            )
         group[name] = cls
         return cls
 
@@ -40,16 +44,19 @@ class ComponentRegistry:
         name: str | None = None,
     ) -> Callable[[type[T]], type[T]] | type[T]:
         """Register a component class directly, or return a decorator that does so."""
-        # Direct mode: first arg is the class itself.
+        # Direct call: register(MyClass) or register(MyClass, "alias").
         if isinstance(cls_or_name, type):
-            return self._do_register(cast(type[T], cls_or_name), name if name is not None else cls_or_name.__name__)
+            cls = cast(type[T], cls_or_name)
+            return self._do_register(cls, name if name is not None else cls.__name__)
 
-        # Decorator mode: first arg is the registration name.
+        # Decorator call: @R.register("alias") — must receive a string name.
         if not isinstance(cls_or_name, str):
             raise TypeError(f"Expected a class or string, got {type(cls_or_name).__name__}")
 
+        registration_name = cls_or_name
+
         def decorator(decorated_cls: type[T]) -> type[T]:
-            return self._do_register(decorated_cls, cls_or_name)
+            return self._do_register(decorated_cls, registration_name)
 
         return decorator
 
diff --git a/reme4/components/file_store/local_file_store.py b/reme4/components/file_store/local_file_store.py
index 171aa66f..c3941373 100644
--- a/reme4/components/file_store/local_file_store.py
+++ b/reme4/components/file_store/local_file_store.py
@@ -111,40 +111,69 @@ class LocalFileStore(BaseFileStore):
         assert self.file_graph is not None
 
         old_map = {n.path: n for n in await self.file_graph.get_nodes([node.path for node, _ in files])}
+        new_nodes, needs_embed, keyword_docs = self._stage_upsert(files, old_map)
 
+        await self.file_graph.upsert_nodes(new_nodes)
+        await self._embed_pending(needs_embed)
+        if self.keyword_index and keyword_docs:
+            await self.keyword_index.add_docs(keyword_docs)
+
+    def _stage_upsert(
+        self,
+        files: list[tuple[FileNode, list[FileChunk]]],
+        old_map: dict[str, FileNode],
+    ) -> tuple[list[FileNode], list[FileChunk], dict[str, str]]:
+        """Mutate self.file_chunks for each file and collect the work the I/O step needs:
+        new graph nodes, chunks still needing an embedding, and keyword docs to index.
+        """
         new_nodes: list[FileNode] = []
         needs_embed: list[FileChunk] = []
         keyword_docs: dict[str, str] = {}
         for node, chunks in files:
-            old_node: FileNode | None = old_map.get(node.path)
-            cached: dict = {}
-            if old_node and self.embedding_model:
-                for cid in old_node.chunk_ids:
-                    old = self.file_chunks.pop(cid, None)
-                    if old and old.embedding is not None:
-                        cached[cid] = old.embedding
-
+            cached = self._evict_prior_chunks(old_map.get(node.path))
             node.chunk_ids = []
             for c in chunks:
-                if self.embedding_model and c.embedding is None:
-                    if c.id in cached:
-                        c.embedding = cached[c.id]
-                    elif c.text:
-                        needs_embed.append(c)
+                self._reuse_or_queue_embedding(c, cached, needs_embed)
                 node.chunk_ids.append(c.id)
                 self.file_chunks[c.id] = c
                 if c.text:
                     keyword_docs[c.id] = c.text
             new_nodes.append(node)
+        return new_nodes, needs_embed, keyword_docs
 
-        await self.file_graph.upsert_nodes(new_nodes)
-        if needs_embed and self.embedding_model:
-            try:
-                await self.embedding_model.get_node_embeddings(needs_embed)
-            except Exception as e:
-                self._disable_embedding(f"upsert: {type(e).__name__}: {e}")
-        if self.keyword_index and keyword_docs:
-            await self.keyword_index.add_docs(keyword_docs)
+    def _evict_prior_chunks(self, old_node: FileNode | None) -> dict[str, np.ndarray]:
+        """Drop chunks for the path being re-upserted; keep their embeddings around so
+        a new chunk reusing the same id avoids a redundant embedding call.
+        """
+        cached: dict[str, np.ndarray] = {}
+        if not (old_node and self.embedding_model):
+            return cached
+        for cid in old_node.chunk_ids:
+            old = self.file_chunks.pop(cid, None)
+            if old and old.embedding is not None:
+                cached[cid] = old.embedding
+        return cached
+
+    def _reuse_or_queue_embedding(
+        self,
+        chunk: FileChunk,
+        cached: dict[str, np.ndarray],
+        needs_embed: list[FileChunk],
+    ) -> None:
+        if not self.embedding_model or chunk.embedding is not None:
+            return
+        if chunk.id in cached:
+            chunk.embedding = cached[chunk.id]
+        elif chunk.text:
+            needs_embed.append(chunk)
+
+    async def _embed_pending(self, chunks: list[FileChunk]) -> None:
+        if not (chunks and self.embedding_model):
+            return
+        try:
+            await self.embedding_model.get_node_embeddings(chunks)
+        except Exception as e:
+            self._disable_embedding(f"upsert: {type(e).__name__}: {e}")
 
     async def delete(self, path: str | list[str]) -> None:
         assert self.file_graph is not None
diff --git a/reme4/components/job/background_job.py b/reme4/components/job/background_job.py
index 12251122..c72de54f 100644
--- a/reme4/components/job/background_job.py
+++ b/reme4/components/job/background_job.py
@@ -35,9 +35,10 @@ class BackgroundJob(BaseJob):
         backoff_cap: float = 60.0,
         close_timeout: float = 5.0,
         attempt_reset_after: float = 60.0,
+        enable_serve: bool = False,
         **kwargs,
     ):
-        super().__init__(**kwargs)
+        super().__init__(enable_serve=enable_serve, **kwargs)
         self.supervisor: bool = supervisor
         self.backoff_base: float = backoff_base
         self.backoff_cap: float = backoff_cap
@@ -54,18 +55,38 @@ class BackgroundJob(BaseJob):
     async def _close(self) -> None:
         if self._stop_event is not None:
             self._stop_event.set()
-        if self._task is not None:
-            try:
-                await asyncio.wait_for(asyncio.shield(self._task), timeout=self.close_timeout)
-            except asyncio.TimeoutError:
-                self._task.cancel()
-                with contextlib.suppress(BaseException):
-                    await self._task
-            except Exception:
-                self.logger.exception(f"Background task '{self.name}' raised during close")
-            self._task = None
+        await self._shutdown_task()
         await super()._close()
 
+    async def _shutdown_task(self) -> None:
+        """Wait close_timeout for graceful exit, then force-cancel."""
+        if self._task is None:
+            return
+        try:
+            # shield prevents wait_for's cancellation from propagating to the task itself,
+            # so a timeout here truly times out instead of cancelling silently.
+            await asyncio.wait_for(asyncio.shield(self._task), timeout=self.close_timeout)
+        except asyncio.TimeoutError:
+            self._task.cancel()
+            with contextlib.suppress(BaseException):
+                await self._task
+        except Exception:
+            self.logger.exception(f"Background task '{self.name}' raised during close")
+        self._task = None
+
+    def _backoff_delay(self, attempt: int) -> float:
+        """Exponential backoff with ±50% jitter, capped at backoff_cap."""
+        capped = min(self.backoff_base * (2**attempt), self.backoff_cap)
+        return min(capped * (0.5 + random.random()), self.backoff_cap)
+
+    async def _wait_or_stop(self, delay: float) -> None:
+        """Sleep up to delay, returning immediately when stop_event is set."""
+        assert self._stop_event is not None
+        try:
+            await asyncio.wait_for(self._stop_event.wait(), timeout=delay)
+        except asyncio.TimeoutError:
+            pass
+
     async def _run_with_supervisor(self) -> None:
         assert self._stop_event is not None
         attempt = 0
@@ -77,16 +98,13 @@ class BackgroundJob(BaseJob):
             except Exception as e:
                 if not self.supervisor:
                     raise
+                # A long-stable run that just crashed restarts fresh rather than at the capped delay.
                 if time.monotonic() - started_at >= self.attempt_reset_after:
                     attempt = 0
-                capped = min(self.backoff_base * (2**attempt), self.backoff_cap)
-                delay = min(capped * (0.5 + random.random()), self.backoff_cap)
+                delay = self._backoff_delay(attempt)
                 self.logger.exception(f"job body crashed, restart in {delay:.2f}s error={e}")
                 attempt += 1
-                try:
-                    await asyncio.wait_for(self._stop_event.wait(), timeout=delay)
-                except asyncio.TimeoutError:
-                    pass
+                await self._wait_or_stop(delay)
 
     async def __call__(self, **kwargs) -> Response:
         """Default body: run steps in order; errors propagate to supervisor."""
diff --git a/reme4/components/job/base_job.py b/reme4/components/job/base_job.py
index 860849f7..3a414bee 100644
--- a/reme4/components/job/base_job.py
+++ b/reme4/components/job/base_job.py
@@ -23,12 +23,14 @@ class BaseJob(BaseComponent):
         description: str = "",
         parameters: dict | None = None,
         steps: list[ComponentConfig | dict] | None = None,
+        enable_serve: bool = True,
         **kwargs,
     ):
         super().__init__(**kwargs)
         self.description = description
         self.parameters = parameters or {}
         self.step_configs = steps or []
+        self.enable_serve = enable_serve
         # Resolved at start: (cls, params) pairs. Steps are re-instantiated per call so they stay
         # stateless across runs and concurrent invocations don't share mutable step state.
         self.step_specs: list[tuple[type["BaseStep"], dict]] = []
diff --git a/reme4/components/job/stream_job.py b/reme4/components/job/stream_job.py
index 193bf5e2..651d5488 100644
--- a/reme4/components/job/stream_job.py
+++ b/reme4/components/job/stream_job.py
@@ -11,11 +11,12 @@ class StreamJob(BaseJob):
     """Job that streams chunks to a queue instead of returning a Response."""
 
     async def __call__(self, **kwargs) -> None:
-        """Execute steps and stream output; errors are sent as ERROR chunks."""
+        """Run steps; emit failures as ERROR chunks, then a terminal DONE marker."""
         context = RuntimeContext(**kwargs)
         try:
             for step in self._build_steps():
                 await step(context)
         except Exception as e:
             await context.add_stream_string(str(e), ChunkEnum.ERROR)
+        # Always emit DONE so consumers can detach even after an error.
         await context.add_stream_done()
diff --git a/reme4/components/keyword_index/base_keyword_index.py b/reme4/components/keyword_index/base_keyword_index.py
index 86433e97..20e7e9da 100644
--- a/reme4/components/keyword_index/base_keyword_index.py
+++ b/reme4/components/keyword_index/base_keyword_index.py
@@ -1,7 +1,4 @@
-"""Abstract base class for keyword index implementations."""
-
 from abc import abstractmethod
-from pathlib import Path
 
 from ..base_component import BaseComponent
 from ..tokenizer import BaseTokenizer
@@ -9,62 +6,48 @@ from ...enumeration import ComponentEnum
 
 
 class BaseKeywordIndex(BaseComponent):
-    """Abstract base class for keyword index implementations."""
+    """关键词索引基类：定义增、删、查、清的统一接口，由具体实现（如 BM25）继承。"""
 
     component_type = ComponentEnum.KEYWORD_INDEX
 
-    def __init__(self, tokenizer: str = "default", index_version: str = "v1", **kwargs):
+    def __init__(self, tokenizer: str = "default", **kwargs):
         super().__init__(**kwargs)
         from ..tokenizer import RegexTokenizer
 
+        # 绑定分词器，未显式指定时回落到 RegexTokenizer
         self.tokenizer = self.bind(tokenizer, BaseTokenizer, default_factory=RegexTokenizer)
-        self.index_version = index_version
-        self.index_path = self.vault_metadata_path / self.component_type.value
-        self.index_path.mkdir(parents=True, exist_ok=True)
+        self.component_metadata_path.mkdir(parents=True, exist_ok=True)
 
     async def _start(self) -> None:
-        """Load existing index from disk if available."""
         await self.load()
 
     async def _close(self) -> None:
-        """Save index to disk on shutdown."""
         await self.dump()
 
-    @property
-    def index_file(self) -> Path:
-        """Return the pickle file path derived from tokenizer name."""
-        if self.tokenizer is None:
-            raise RuntimeError("Tokenizer not initialized. Call start() first.")
-        name = type(self.tokenizer).__name__.replace("Tokenizer", "").lower()
-        return self.index_path / f"bm25_{name}_{self.index_version}.pkl"
-
     def _tokenize(self, text: str) -> list[str]:
-        """Tokenize a text string into tokens."""
+        """对单段文本调用分词器，返回 token 列表。"""
         if self.tokenizer is None:
             raise RuntimeError("Tokenizer not initialized. Call start() first.")
         return self.tokenizer.tokenize([text])[0]
 
     @abstractmethod
-    async def add_docs(self, docs_dict: dict[str, str]) -> None:
-        """Index or update documents. Mapping of doc_id to content."""
+    async def add_docs(self, docs_dict: dict[str, str]) -> None: ...
 
     @abstractmethod
-    async def delete_docs(self, doc_ids: list[str]) -> None:
-        """Remove documents by their IDs."""
+    async def delete_docs(self, doc_ids: list[str]) -> None: ...
 
     @abstractmethod
-    async def retrieve(self, query: str, limit: int = 3) -> dict[str, float]:
-        """Search documents. Returns {doc_id: score} sorted descending."""
+    async def retrieve(self, query: str, limit: int = 3) -> dict[str, float]: ...
 
     @abstractmethod
-    async def clear(self) -> None:
-        """Reset index to empty state."""
+    async def clear(self) -> None: ...
 
     async def reset_index(self, docs_dict: dict[str, str]) -> None:
-        """Clear index, re-add all documents, and persist."""
+        """清空索引后重新构建，并立即落盘。"""
         await self.clear()
         await self.add_docs(docs_dict)
         await self.dump()
 
     async def optimize_index(self) -> None:
-        """Optimize index for performance. Override in subclass if needed."""
+        """对索引进行物理压缩或重建；基类默认无操作，由子类按需重载。"""
+        pass
diff --git a/reme4/components/keyword_index/bm25_index.py b/reme4/components/keyword_index/bm25_index.py
index 31d978a9..d5058f0e 100644
--- a/reme4/components/keyword_index/bm25_index.py
+++ b/reme4/components/keyword_index/bm25_index.py
@@ -1,27 +1,24 @@
-"""BM25 search engine with persistent index support.
+"""基于 BM25 的倒排索引实现，支持持久化。
 
-Implements Okapi BM25 ranking with a numpy-vectorized inverted index for
-efficient document lookup, incremental updates, and pickle-based persistence.
-
-Storage layout (source of truth):
+核心存储结构（落盘真相源）：
     vocab               : dict[token, token_id]
-    _doc_ids            : list[doc_id] indexed by doc_idx
+    _doc_ids            : list[doc_id]，按 doc_idx 索引
     _doc_id_to_idx      : dict[doc_id, doc_idx]
-    _doc_lens           : np.ndarray[int32] indexed by doc_idx
-    _deleted            : np.ndarray[bool]  indexed by doc_idx (lazy deletion)
-    _doc_token_ids      : list[np.ndarray[int32]] indexed by doc_idx (unique tids per doc)
-    _posting_doc_idxs   : dict[token_id, np.ndarray[int32]]  posting list doc_idxs
-    _posting_tfs        : dict[token_id, np.ndarray[int32]]  posting list tfs (parallel)
+    _doc_lens           : np.ndarray[int32]，按 doc_idx 索引
+    _deleted            : np.ndarray[bool]，按 doc_idx 索引（懒删除标记）
+    _doc_token_ids      : list[np.ndarray[int32]]，每篇文档去重后的 token_id
+    _posting_doc_idxs   : dict[token_id, np.ndarray[int32]]，倒排表的 doc_idx
+    _posting_tfs        : dict[token_id, np.ndarray[int32]]，与上方一一对应的词频
 
-Deletion is lazy: ``_remove_doc`` only flips ``_deleted[idx]``; posting entries
-pointing at the dead idx are masked at query time and physically dropped by
-``optimize_index``. Updating an existing doc_id marks the old slot deleted and
-allocates a fresh idx for the new content.
+删除采用懒标记：_deleted[idx] = True 即视为删除，倒排表中的物理回收由
+optimize_index 统一完成；更新已存在的 doc_id 时，先把旧槽位标记删除，再
+分配新的 idx。
 """
 
 import math
 import pickle
 from collections import Counter
+from pathlib import Path
 
 import numpy as np
 
@@ -31,82 +28,77 @@ from ..component_registry import R
 
 @R.register("bm25")
 class BM25Index(BaseKeywordIndex):
-    """BM25 search engine with numpy-vectorized scoring and file-based persistence.
 
-    Args:
-        k1: Term frequency saturation parameter (default 1.5).
-        b: Document length normalization parameter (default 0.75).
-    """
-
-    def __init__(self, k1: float = 1.5, b: float = 0.75, **kwargs):
+    def __init__(self, k1: float = 1.5, b: float = 0.75, index_version: str = "v1", **kwargs):
         super().__init__(**kwargs)
         self.k1 = k1
         self.b = b
+        self.index_version = index_version
+
+        # 词表与文档元数据
         self.vocab: dict[str, int] = {}
         self._doc_ids: list[str] = []
         self._doc_id_to_idx: dict[str, int] = {}
         self._doc_lens: np.ndarray = np.zeros(0, dtype=np.int32)
         self._deleted: np.ndarray = np.zeros(0, dtype=bool)
         self._doc_token_ids: list[np.ndarray] = []
+
+        # 倒排表：token_id -> (doc_idxs, tfs)
         self._posting_doc_idxs: dict[int, np.ndarray] = {}
         self._posting_tfs: dict[int, np.ndarray] = {}
+
+        # IDF 缓存，对增删与重建索引时失效
         self._idf_cache: dict[int, float] = {}
 
     # -- Properties -----------------------------------------------------------
 
+    @property
+    def index_file(self) -> Path:
+        """落盘文件路径，包含分词器名与索引版本，便于区分不同配置。"""
+        if self.tokenizer is None:
+            raise RuntimeError("Tokenizer not initialized. Call start() first.")
+        name = type(self.tokenizer).__name__.replace("Tokenizer", "").lower()
+        return self.component_metadata_path / f"bm25_{name}_{self.index_version}.pkl"
+
     @property
     def n_docs(self) -> int:
-        """Number of indexed (non-deleted) documents."""
-        if self._deleted.size == 0:
-            return 0
-        return int((~self._deleted).sum())
+        """当前存活文档数（排除懒删除）。"""
+        return 0 if self._deleted.size == 0 else int((~self._deleted).sum())
 
     @property
     def total_len(self) -> int:
-        """Total tokens across non-deleted documents."""
-        if self._deleted.size == 0:
-            return 0
-        return int(self._doc_lens[~self._deleted].sum())
+        """所有存活文档的 token 总数。"""
+        return 0 if self._deleted.size == 0 else int(self._doc_lens[~self._deleted].sum())
 
     @property
     def avg_len(self) -> float:
-        """Average document length in tokens (non-deleted only)."""
+        """存活文档的平均长度，用于 BM25 长度归一化。"""
         n = self.n_docs
         return self.total_len / n if n > 0 else 0.0
 
     @property
     def doc_meta(self) -> dict[str, dict]:
-        """Dict-view of {doc_id: {"len", "token_ids"}} for non-deleted docs.
-
-        Built on demand from the numpy-backed storage; kept for backward
-        compatibility with callers (and tests) that read this shape.
-        """
-        out: dict[str, dict] = {}
-        for idx, doc_id in enumerate(self._doc_ids):
-            if self._deleted[idx]:
-                continue
-            out[doc_id] = {
+        """对外暴露每篇存活文档的长度与去重后的 token_id 集合。"""
+        return {
+            self._doc_ids[idx]: {
                 "len": int(self._doc_lens[idx]),
                 "token_ids": {int(t) for t in self._doc_token_ids[idx]},
             }
-        return out
+            for idx in range(len(self._doc_ids))
+            if not self._deleted[idx]
+        }
 
     @property
     def inverted_index(self) -> dict[int, dict[str, int]]:
-        """Dict-view of {token_id: {doc_id: tf}} excluding deleted docs.
-
-        Built on demand from the numpy-backed storage; kept for backward
-        compatibility. Empty posting lists (all entries deleted) are omitted.
-        """
+        """重建可读形式的倒排表：token_id -> {doc_id: tf}，跳过已删除文档。"""
         out: dict[int, dict[str, int]] = {}
         for tid, doc_idxs in self._posting_doc_idxs.items():
             tfs = self._posting_tfs[tid]
-            posting: dict[str, int] = {}
-            for i, tf in zip(doc_idxs, tfs):
-                i = int(i)
-                if self._deleted[i]:
-                    continue
-                posting[self._doc_ids[i]] = int(tf)
+            posting = {
+                self._doc_ids[int(i)]: int(tf)
+                for i, tf in zip(doc_idxs, tfs)
+                if not self._deleted[int(i)]
+            }
             if posting:
                 out[tid] = posting
         return out
@@ -114,9 +106,9 @@ class BM25Index(BaseKeywordIndex):
     # -- Internal helpers -----------------------------------------------------
 
     def _tokens_to_ids(self, tokens: list[str]) -> list[int]:
-        """Map tokens to integer IDs, assigning new IDs on first encounter."""
+        """将 token 转为 id；遇到新词时自动分配新的 token_id。"""
         vocab = self.vocab
-        ids = []
+        ids: list[int] = []
         for token in tokens:
             token = token.strip()
             if not token:
@@ -129,7 +121,7 @@ class BM25Index(BaseKeywordIndex):
         return ids
 
     def _remove_doc(self, doc_id: str) -> None:
-        """Mark a document deleted. Posting cleanup deferred to ``optimize_index``."""
+        """懒删除：仅置 _deleted 位并解除 doc_id 映射，不动倒排表。"""
         idx = self._doc_id_to_idx.get(doc_id)
         if idx is None or self._deleted[idx]:
             return
@@ -138,7 +130,7 @@ class BM25Index(BaseKeywordIndex):
         self._idf_cache = {}
 
     def _get_idf(self, token_id: int, n_docs: int | None = None) -> float:
-        """Compute and cache IDF for a token ID against current active doc set."""
+        """计算并缓存 token 的 IDF；存活文档数发生变化时缓存会被清空。"""
         if token_id in self._idf_cache:
             return self._idf_cache[token_id]
         doc_idxs = self._posting_doc_idxs.get(token_id)
@@ -146,68 +138,36 @@ class BM25Index(BaseKeywordIndex):
             self._idf_cache[token_id] = 0.0
             return 0.0
         df = int((~self._deleted[doc_idxs]).sum())
-        if df == 0:
-            self._idf_cache[token_id] = 0.0
-            return 0.0
         if n_docs is None:
             n_docs = self.n_docs
-        self._idf_cache[token_id] = math.log(1 + (n_docs - df + 0.5) / (df + 0.5))
-        return self._idf_cache[token_id]
+        idf = math.log(1 + (n_docs - df + 0.5) / (df + 0.5)) if df else 0.0
+        self._idf_cache[token_id] = idf
+        return idf
 
-    # -- Public API -----------------------------------------------------------
+    def _prepare_doc(self, doc_id: str, content: str) -> tuple[np.ndarray, int, Counter] | None:
+        """分词并统计词频；若 doc_id 已存在则先标记旧版为删除。空文档返回 None。"""
+        self._remove_doc(doc_id)
+        token_ids = self._tokens_to_ids(self._tokenize(content))
+        if not token_ids:
+            return None
+        counts = Counter(token_ids)
+        unique_tids = np.fromiter(counts.keys(), dtype=np.int32, count=len(counts))
+        return unique_tids, len(token_ids), counts
 
-    async def add_docs(self, docs_dict: dict[str, str]) -> None:
-        """Index or update multiple documents. Mapping of doc_id to content.
-
-        Updating an existing doc_id marks the old slot deleted and allocates
-        a new doc_idx, so the next ``optimize_index`` reclaims its postings.
-        """
-        if not docs_dict:
+    def _append_doc_arrays(
+        self, new_doc_ids: list[str], new_doc_lens: list[int], new_doc_token_ids: list[np.ndarray]
+    ) -> None:
+        """把一批新文档的元数据一次性追加到文档数组中。"""
+        if not new_doc_ids:
             return
+        self._doc_ids.extend(new_doc_ids)
+        self._doc_token_ids.extend(new_doc_token_ids)
+        self._doc_lens = np.concatenate([self._doc_lens, np.array(new_doc_lens, dtype=np.int32)])
+        self._deleted = np.concatenate([self._deleted, np.zeros(len(new_doc_ids), dtype=bool)])
 
-        new_doc_ids: list[str] = []
-        new_doc_lens: list[int] = []
-        new_doc_token_ids: list[np.ndarray] = []
-        pending_postings: dict[int, list[tuple[int, int]]] = {}
-
-        next_idx = len(self._doc_ids)
-
-        for doc_id, content in docs_dict.items():
-            old_idx = self._doc_id_to_idx.get(doc_id)
-            if old_idx is not None and not self._deleted[old_idx]:
-                self._deleted[old_idx] = True
-                self._doc_id_to_idx.pop(doc_id, None)
-
-            token_ids = self._tokens_to_ids(self._tokenize(content))
-            if not token_ids:
-                continue
-
-            token_counts = Counter(token_ids)
-            unique_tids = np.fromiter(
-                token_counts.keys(), dtype=np.int32, count=len(token_counts)
-            )
-
-            idx = next_idx
-            next_idx += 1
-            new_doc_ids.append(doc_id)
-            new_doc_lens.append(len(token_ids))
-            new_doc_token_ids.append(unique_tids)
-            self._doc_id_to_idx[doc_id] = idx
-
-            for tid, tf in token_counts.items():
-                pending_postings.setdefault(tid, []).append((idx, tf))
-
-        if new_doc_ids:
-            self._doc_ids.extend(new_doc_ids)
-            self._doc_token_ids.extend(new_doc_token_ids)
-            self._doc_lens = np.concatenate(
-                [self._doc_lens, np.array(new_doc_lens, dtype=np.int32)]
-            )
-            self._deleted = np.concatenate(
-                [self._deleted, np.zeros(len(new_doc_ids), dtype=bool)]
-            )
-
-        for tid, items in pending_postings.items():
+    def _extend_postings(self, pending: dict[int, list[tuple[int, int]]]) -> None:
+        """把待写入的 (doc_idx, tf) 增量按 token 追加到倒排表。"""
+        for tid, items in pending.items():
             n = len(items)
             new_idxs = np.fromiter((idx for idx, _ in items), dtype=np.int32, count=n)
             new_tfs = np.fromiter((tf for _, tf in items), dtype=np.int32, count=n)
@@ -218,113 +178,150 @@ class BM25Index(BaseKeywordIndex):
                 self._posting_doc_idxs[tid] = new_idxs
                 self._posting_tfs[tid] = new_tfs
 
+    def _encode_query(self, query: str) -> list[int]:
+        """切词、过滤未登录词并去重，返回查询的 token_id 列表。"""
+        vocab = self.vocab
+        return list(dict.fromkeys(vocab[t] for t in self._tokenize(query) if t in vocab))
+
+    def _top_k(self, scores: np.ndarray, limit: int) -> np.ndarray:
+        """挑出得分前 limit 名（且严格大于 0）的索引，按得分降序排列。"""
+        if limit <= 0:
+            return np.empty(0, dtype=np.int64)
+        positive_count = int((scores > 0).sum())
+        if positive_count == 0:
+            return np.empty(0, dtype=np.int64)
+        k = min(limit, positive_count)
+        if k >= scores.size:
+            return np.argsort(-scores)[:k]
+        top = np.argpartition(-scores, k - 1)[:k]
+        return top[np.argsort(-scores[top])]
+
+    # -- Public API -----------------------------------------------------------
+
+    async def add_docs(self, docs_dict: dict[str, str]) -> None:
+        """批量加入文档；已存在的 doc_id 会被替换为新版本。"""
+        if not docs_dict:
+            return
+
+        new_doc_ids: list[str] = []
+        new_doc_lens: list[int] = []
+        new_doc_token_ids: list[np.ndarray] = []
+        pending: dict[int, list[tuple[int, int]]] = {}
+        next_idx = len(self._doc_ids)
+
+        for doc_id, content in docs_dict.items():
+            prepared = self._prepare_doc(doc_id, content)
+            if prepared is None:
+                continue
+            unique_tids, n_tokens, token_counts = prepared
+
+            idx = next_idx
+            next_idx += 1
+            new_doc_ids.append(doc_id)
+            new_doc_lens.append(n_tokens)
+            new_doc_token_ids.append(unique_tids)
+            self._doc_id_to_idx[doc_id] = idx
+            for tid, tf in token_counts.items():
+                pending.setdefault(tid, []).append((idx, tf))
+
+        self._append_doc_arrays(new_doc_ids, new_doc_lens, new_doc_token_ids)
+        self._extend_postings(pending)
         self._idf_cache = {}
 
     async def delete_docs(self, doc_ids: list[str]) -> None:
-        """Remove documents by their IDs."""
+        """批量懒删除；倒排表中的物理回收由 optimize_index 完成。"""
         for doc_id in doc_ids:
             self._remove_doc(doc_id)
         self._idf_cache = {}
 
-    async def retrieve(self, query: str, limit: int = 3) -> dict[str, float]:
-        """Search documents. Returns {doc_id: score} sorted descending."""
-        n_slots = self._doc_lens.size
-        if n_slots == 0:
-            return {}
-
-        vocab = self.vocab
-        query_ids = list(dict.fromkeys(vocab[t] for t in self._tokenize(query) if t in vocab))
-        if not query_ids:
-            return {}
-
-        n_docs = self.n_docs
-        if n_docs == 0:
-            return {}
-
+    def _score_query(self, query_ids: list[int], n_docs: int) -> np.ndarray:
+        """对所有文档计算 BM25 得分，已删除文档置 0。"""
         avg_len = self.total_len / n_docs
         k1, b = self.k1, self.b
         denom_base = k1 * (1.0 - b)
         denom_norm = k1 * b / avg_len if avg_len > 0 else 0.0
 
-        scores = np.zeros(n_slots, dtype=np.float32)
-
+        scores = np.zeros(self._doc_lens.size, dtype=np.float32)
         for tid in query_ids:
             doc_idxs = self._posting_doc_idxs.get(tid)
             if doc_idxs is None or doc_idxs.size == 0:
                 continue
-            idf = self._get_idf(tid, n_docs=n_docs)
+            idf = self._get_idf(tid, n_docs)
             if idf == 0.0:
                 continue
             tfs = self._posting_tfs[tid].astype(np.float32)
             d_lens = self._doc_lens[doc_idxs].astype(np.float32)
-            tf_score = tfs * (k1 + 1.0) / (tfs + denom_base + denom_norm * d_lens)
-            # Each doc_idx appears at most once per posting list (Counter dedups
-            # within a doc, and updates allocate a fresh idx), so direct
-            # advanced-indexing assignment-add is safe.
-            scores[doc_idxs] += idf * tf_score
+            # 同一倒排表中每个 doc_idx 至多出现一次：Counter 已在文档内去重，
+            # 文档更新也会分配新的 idx，因此可安全使用花式索引累加。
+            scores[doc_idxs] += idf * tfs * (k1 + 1.0) / (tfs + denom_base + denom_norm * d_lens)
 
         if self._deleted.any():
             scores[self._deleted] = 0.0
+        return scores
 
-        positive_count = int((scores > 0).sum())
-        if positive_count == 0:
+    async def retrieve(self, query: str, limit: int = 3) -> dict[str, float]:
+        """对查询做 BM25 召回，返回 {doc_id: score}，按得分降序。"""
+        n_docs = self.n_docs
+        if n_docs == 0:
+            return {}
+        query_ids = self._encode_query(query)
+        if not query_ids:
             return {}
-        k = min(limit, positive_count)
-        if k >= n_slots:
-            top_idxs = np.argsort(-scores)[:k]
-        else:
-            top_idxs = np.argpartition(-scores, k - 1)[:k]
-            top_idxs = top_idxs[np.argsort(-scores[top_idxs])]
 
+        scores = self._score_query(query_ids, n_docs)
+        top_idxs = self._top_k(scores, limit)
+        return {self._doc_ids[int(i)]: float(scores[int(i)]) for i in top_idxs}
+
+    # -- Persistence ----------------------------------------------------------
+
+    def _snapshot(self) -> dict:
+        """收集需要落盘的全部字段，集中在一处以便与 _restore 对齐。"""
         return {
-            self._doc_ids[int(i)]: float(scores[int(i)])
-            for i in top_idxs
-            if scores[int(i)] > 0
+            "vocab": self.vocab,
+            "doc_ids": self._doc_ids,
+            "doc_id_to_idx": self._doc_id_to_idx,
+            "doc_lens": self._doc_lens,
+            "deleted": self._deleted,
+            "doc_token_ids": self._doc_token_ids,
+            "posting_doc_idxs": self._posting_doc_idxs,
+            "posting_tfs": self._posting_tfs,
+            "k1": self.k1,
+            "b": self.b,
         }
 
+    def _restore(self, data: dict) -> None:
+        """从 _snapshot 产生的字典还原索引内部状态。"""
+        self.vocab = data["vocab"]
+        self._doc_ids = data["doc_ids"]
+        self._doc_id_to_idx = data["doc_id_to_idx"]
+        self._doc_lens = data["doc_lens"]
+        self._deleted = data["deleted"]
+        self._doc_token_ids = data["doc_token_ids"]
+        self._posting_doc_idxs = data["posting_doc_idxs"]
+        self._posting_tfs = data["posting_tfs"]
+        self.k1 = data.get("k1", 1.5)
+        self.b = data.get("b", 0.75)
+        self._idf_cache = {}
+
     async def dump(self) -> None:
-        """Persist index to disk via pickle (atomic rename)."""
+        """通过临时文件 + 原子替换的方式持久化索引，避免半写状态。"""
         try:
             tmp = self.index_file.with_suffix(".tmp")
             with open(tmp, "wb") as f:
-                pickle.dump(
-                    {
-                        "vocab": self.vocab,
-                        "doc_ids": self._doc_ids,
-                        "doc_id_to_idx": self._doc_id_to_idx,
-                        "doc_lens": self._doc_lens,
-                        "deleted": self._deleted,
-                        "doc_token_ids": self._doc_token_ids,
-                        "posting_doc_idxs": self._posting_doc_idxs,
-                        "posting_tfs": self._posting_tfs,
-                        "k1": self.k1,
-                        "b": self.b,
-                    },
-                    f,
-                )
+                pickle.dump(self._snapshot(), f)
             tmp.replace(self.index_file)
             self.logger.info(f"Saved {self.n_docs} docs to {self.index_file}")
         except Exception as e:
             self.logger.exception(f"Failed to write {self.index_file}: {e}")
 
     async def load(self) -> None:
-        """Load index from disk. No-op if file missing; clears index on corruption."""
+        """读取持久化文件并还原索引；文件不存在则不做事，损坏则清空。"""
         if not self.index_file.exists():
             return
         try:
             with open(self.index_file, "rb") as f:
                 data = pickle.load(f)
-            self.vocab = data["vocab"]
-            self._doc_ids = data["doc_ids"]
-            self._doc_id_to_idx = data["doc_id_to_idx"]
-            self._doc_lens = data["doc_lens"]
-            self._deleted = data["deleted"]
-            self._doc_token_ids = data["doc_token_ids"]
-            self._posting_doc_idxs = data["posting_doc_idxs"]
-            self._posting_tfs = data["posting_tfs"]
-            self.k1 = data.get("k1", 1.5)
-            self.b = data.get("b", 0.75)
-            self._idf_cache = {}
+            self._restore(data)
             self.logger.info(f"Loaded {self.n_docs} docs from {self.index_file}")
         except Exception as e:
             self.logger.exception(f"Failed to load index: {e}")
@@ -332,7 +329,7 @@ class BM25Index(BaseKeywordIndex):
             await self.clear()
 
     async def clear(self) -> None:
-        """Reset index to empty state and remove persisted file."""
+        """清空内存中的索引并删除持久化文件。"""
         self.vocab = {}
         self._doc_ids = []
         self._doc_id_to_idx = {}
@@ -344,64 +341,86 @@ class BM25Index(BaseKeywordIndex):
         self._idf_cache = {}
         self.index_file.unlink(missing_ok=True)
 
+    # -- Compaction -----------------------------------------------------------
+
+    def _build_idx_remap(self, active_mask: np.ndarray) -> tuple[np.ndarray, int]:
+        """构造 old_idx → new_idx 的映射数组（被删槽位为 -1），并返回存活数量。"""
+        active_old_idxs = np.where(active_mask)[0]
+        n_active = int(active_old_idxs.size)
+        remap = -np.ones(self._deleted.size, dtype=np.int32)
+        remap[active_old_idxs] = np.arange(n_active, dtype=np.int32)
+        return remap, n_active
+
+    def _compact_vocab(self, active_mask: np.ndarray) -> tuple[dict[str, int], dict[int, int]]:
+        """只保留仍被任意存活文档引用的 token，重排成连续的新 token_id。"""
+        used_tids = {
+            tid for tid, doc_idxs in self._posting_doc_idxs.items()
+            if active_mask[doc_idxs].any()
+        }
+        new_vocab: dict[str, int] = {}
+        old_to_new: dict[int, int] = {}
+        for token, old_tid in self.vocab.items():
+            if old_tid in used_tids:
+                new_tid = len(new_vocab)
+                new_vocab[token] = new_tid
+                old_to_new[old_tid] = new_tid
+        return new_vocab, old_to_new
+
+    def _compact_postings(
+        self,
+        active_mask: np.ndarray,
+        old_to_new_idx: np.ndarray,
+        old_tid_to_new: dict[int, int],
+    ) -> tuple[dict[int, np.ndarray], dict[int, np.ndarray]]:
+        """剔除删除文档并按新 idx/tid 重写倒排表。"""
+        new_idxs: dict[int, np.ndarray] = {}
+        new_tfs: dict[int, np.ndarray] = {}
+        for tid, doc_idxs in self._posting_doc_idxs.items():
+            if tid not in old_tid_to_new:
+                continue
+            mask = active_mask[doc_idxs]
+            new_tid = old_tid_to_new[tid]
+            new_idxs[new_tid] = old_to_new_idx[doc_idxs[mask]].astype(np.int32, copy=False)
+            new_tfs[new_tid] = self._posting_tfs[tid][mask].astype(np.int32, copy=False)
+        return new_idxs, new_tfs
+
+    def _compact_docs(
+        self, active_mask: np.ndarray, old_tid_to_new: dict[int, int]
+    ) -> tuple[list[str], list[np.ndarray]]:
+        """在压缩后的词表下重建存活文档的 doc_id 列表与去重 token_id 数组。"""
+        active_old_idxs = np.where(active_mask)[0]
+        new_doc_ids = [self._doc_ids[int(i)] for i in active_old_idxs]
+        new_doc_token_ids = [
+            np.fromiter(
+                (old_tid_to_new[int(t)] for t in self._doc_token_ids[int(i)] if int(t) in old_tid_to_new),
+                dtype=np.int32,
+            )
+            for i in active_old_idxs
+        ]
+        return new_doc_ids, new_doc_token_ids
+
     async def optimize_index(self) -> None:
-        """Compact: drop deleted docs, reassign doc_idx, prune unused tokens."""
+        """物理回收懒删除的文档与未被引用的词表项，重建紧凑索引。"""
         if self._deleted.size == 0:
             return
-
         active_mask = ~self._deleted
         if not active_mask.any():
             await self.clear()
             return
 
-        active_old_idxs = np.where(active_mask)[0]
-        n_active = int(active_old_idxs.size)
-        old_to_new_idx = -np.ones(self._deleted.size, dtype=np.int32)
-        old_to_new_idx[active_old_idxs] = np.arange(n_active, dtype=np.int32)
-
-        new_doc_ids = [self._doc_ids[int(i)] for i in active_old_idxs]
-        new_doc_lens = self._doc_lens[active_mask].astype(np.int32, copy=True)
-        new_doc_token_ids_pre = [self._doc_token_ids[int(i)] for i in active_old_idxs]
-        new_doc_id_to_idx = {doc_id: i for i, doc_id in enumerate(new_doc_ids)}
-
-        used_tids: set[int] = set()
-        for tid, doc_idxs in self._posting_doc_idxs.items():
-            if active_mask[doc_idxs].any():
-                used_tids.add(tid)
-
-        old_tid_to_new: dict[int, int] = {}
-        new_vocab: dict[str, int] = {}
-        for token, old_tid in self.vocab.items():
-            if old_tid in used_tids:
-                new_tid = len(new_vocab)
-                new_vocab[token] = new_tid
-                old_tid_to_new[old_tid] = new_tid
-
-        new_posting_doc_idxs: dict[int, np.ndarray] = {}
-        new_posting_tfs: dict[int, np.ndarray] = {}
-        for tid, doc_idxs in self._posting_doc_idxs.items():
-            if tid not in old_tid_to_new:
-                continue
-            mask = active_mask[doc_idxs]
-            kept_idxs = old_to_new_idx[doc_idxs[mask]].astype(np.int32, copy=False)
-            kept_tfs = self._posting_tfs[tid][mask].astype(np.int32, copy=False)
-            new_posting_doc_idxs[old_tid_to_new[tid]] = kept_idxs
-            new_posting_tfs[old_tid_to_new[tid]] = kept_tfs
-
-        new_doc_token_ids: list[np.ndarray] = []
-        for arr in new_doc_token_ids_pre:
-            remapped = np.fromiter(
-                (old_tid_to_new[int(t)] for t in arr if int(t) in old_tid_to_new),
-                dtype=np.int32,
-            )
-            new_doc_token_ids.append(remapped)
+        old_to_new_idx, n_active = self._build_idx_remap(active_mask)
+        new_vocab, old_tid_to_new = self._compact_vocab(active_mask)
+        new_posting_idxs, new_posting_tfs = self._compact_postings(
+            active_mask, old_to_new_idx, old_tid_to_new
+        )
+        new_doc_ids, new_doc_token_ids = self._compact_docs(active_mask, old_tid_to_new)
 
         self.vocab = new_vocab
         self._doc_ids = new_doc_ids
-        self._doc_id_to_idx = new_doc_id_to_idx
-        self._doc_lens = new_doc_lens
+        self._doc_id_to_idx = {doc_id: i for i, doc_id in enumerate(new_doc_ids)}
+        self._doc_lens = self._doc_lens[active_mask].astype(np.int32, copy=True)
         self._deleted = np.zeros(n_active, dtype=bool)
         self._doc_token_ids = new_doc_token_ids
-        self._posting_doc_idxs = new_posting_doc_idxs
+        self._posting_doc_idxs = new_posting_idxs
         self._posting_tfs = new_posting_tfs
         self._idf_cache = {}
diff --git a/reme4/components/prompt_handler.py b/reme4/components/prompt_handler.py
index a9ae203a..db97a963 100644
--- a/reme4/components/prompt_handler.py
+++ b/reme4/components/prompt_handler.py
@@ -8,26 +8,27 @@ from string import Formatter
 
 import yaml
 
-# Matches a leading flag tag like "[verbose] some text".
+# Matches a leading flag tag at line start: "[flag] rest of line".
 _FLAG_PATTERN = re.compile(r"^\[(\w+)]")
 
 
 class PromptHandler:
-    """Loads prompts from YAML/JSON or class-adjacent files and formats them.
+    """Loads prompts from YAML/JSON files and renders them with optional flags.
 
-    Templates may carry a language suffix (``key_en``, ``key_zh``); ``get_prompt``
-    falls back to the bare key when no localized variant exists. ``prompt_format``
-    additionally supports per-line flags such as ``[verbose] extra text`` that
-    are kept only when the matching flag kwarg is truthy.
+    Template keys may carry a language suffix (``key_en``, ``key_zh``); lookups
+    fall back to the bare key when no localized variant exists. Lines tagged
+    with ``[flag]`` are kept only when the matching boolean kwarg is truthy.
     """
 
     _SUPPORTED_EXTENSIONS = {".yaml", ".yml", ".json"}
 
     def __init__(self, language: str = "", **kwargs):
-        # Only string entries are treated as prompts; other kwargs are ignored.
+        # Non-string kwargs are silently dropped — prompts must be strings.
         self.data: dict[str, str] = {k: v for k, v in kwargs.items() if isinstance(v, str)}
         self.language: str = language.strip()
 
+    # ----- Loading -------------------------------------------------------
+
     def load_prompt_by_file(
         self,
         prompt_file_path: str | Path | None = None,
@@ -41,13 +42,18 @@ class PromptHandler:
         if not path.exists() or path.suffix.lower() not in self._SUPPORTED_EXTENSIONS:
             return self
 
+        return self.load_prompt_dict(self._parse_prompt_file(path), overwrite)
+
+    @staticmethod
+    def _parse_prompt_file(path: Path) -> dict | None:
+        """Parse a YAML or JSON prompt file; return None on any parse error."""
         try:
             with path.open(encoding="utf-8") as f:
-                prompt_dict = yaml.safe_load(f) if path.suffix.lower() in (".yaml", ".yml") else json.load(f)
+                if path.suffix.lower() in (".yaml", ".yml"):
+                    return yaml.safe_load(f)
+                return json.load(f)
         except (json.JSONDecodeError, yaml.YAMLError, OSError):
-            return self
-
-        return self.load_prompt_dict(prompt_dict, overwrite)
+            return None
 
     def load_prompt_by_class(self, cls: type, overwrite: bool = True) -> "PromptHandler":
         """Load prompts from ``<class_module>.yaml`` (or ``.yml``) next to `cls`."""
@@ -57,9 +63,9 @@ class PromptHandler:
             return self
 
         for ext in (".yaml", ".yml"):
-            if (prompt_path := base_path.with_suffix(ext)).exists():
-                return self.load_prompt_by_file(prompt_path, overwrite)
-
+            candidate = base_path.with_suffix(ext)
+            if candidate.exists():
+                return self.load_prompt_by_file(candidate, overwrite)
         return self
 
     def load_prompt_dict(self, prompt_dict: dict | None = None, overwrite: bool = True) -> "PromptHandler":
@@ -70,21 +76,28 @@ class PromptHandler:
         for key, value in prompt_dict.items():
             if isinstance(value, str) and (overwrite or key not in self.data):
                 self.data[key] = value
-
         return self
 
+    # ----- Lookup --------------------------------------------------------
+
+    def _candidate_keys(self, prompt_name: str) -> tuple[str, ...]:
+        """Lookup order: localized key first when a language is set, then bare key."""
+        if self.language:
+            return (f"{prompt_name}_{self.language}", prompt_name)
+        return (prompt_name,)
+
     def get_prompt(self, prompt_name: str) -> str:
         """Return the template, preferring the language-suffixed variant when set."""
-        for key in (f"{prompt_name}_{self.language}", prompt_name) if self.language else (prompt_name,):
+        for key in self._candidate_keys(prompt_name):
             if key in self.data:
                 return self.data[key].strip()
-
-        raise KeyError(f"Prompt '{prompt_name}' not found. Available: {list(self.data.keys())[:10]}")
+        raise KeyError(
+            f"Prompt '{prompt_name}' not found. Available: {list(self.data.keys())[:10]}"
+        )
 
     def has_prompt(self, prompt_name: str) -> bool:
         """True if either the localized or bare prompt is registered."""
-        keys = (f"{prompt_name}_{self.language}", prompt_name) if self.language else (prompt_name,)
-        return any(k in self.data for k in keys)
+        return any(k in self.data for k in self._candidate_keys(prompt_name))
 
     def list_prompts(self, language_filter: str | None = None) -> list[str]:
         """List all keys, optionally filtered to those ending with ``_<language>``."""
@@ -93,33 +106,44 @@ class PromptHandler:
         suffix = f"_{language_filter.strip()}"
         return [k for k in self.data if k.endswith(suffix)]
 
-    def prompt_format(self, prompt_name: str, validate: bool = True, **kwargs) -> str:
-        """Render a prompt: strip inactive flag-lines, then ``str.format`` it.
+    # ----- Formatting ----------------------------------------------------
 
-        Boolean kwargs are treated as flags controlling ``[flag]`` line filtering.
-        Remaining kwargs become positional substitutions for ``{var}`` placeholders.
-        With `validate=True`, missing substitutions raise ``ValueError``.
+    def prompt_format(self, prompt_name: str, validate: bool = True, **kwargs) -> str:
+        """Render a prompt: strip inactive ``[flag]`` lines, then ``str.format`` it.
+
+        Boolean kwargs are treated as flag toggles; the rest are format variables.
+        With ``validate=True``, any missing ``{var}`` placeholder raises ``ValueError``.
         """
         prompt = self.get_prompt(prompt_name)
         flags = {k: v for k, v in kwargs.items() if isinstance(v, bool)}
         formats = {k: v for k, v in kwargs.items() if not isinstance(v, bool)}
 
-        # Keep lines without flags; otherwise keep when at least one flag is enabled.
         if flags:
-            lines = []
-            for line in prompt.split("\n"):
-                active_flags = _FLAG_PATTERN.findall(line)
-                cleaned = _FLAG_PATTERN.sub("", line).lstrip()
-                if not active_flags or any(flags.get(f, False) for f in active_flags):
-                    lines.append(cleaned)
-            prompt = "\n".join(lines)
-
+            prompt = self._apply_flag_filter(prompt, flags)
         if validate:
-            required = {f for _, f, _, _ in Formatter().parse(prompt) if f is not None}
-            if missing := required - set(formats.keys()):
-                raise ValueError(f"Missing format variables for '{prompt_name}': {sorted(missing)}")
+            self._check_required_vars(prompt, formats, prompt_name)
 
         return prompt.format(**formats).strip() if formats else prompt
 
+    @staticmethod
+    def _apply_flag_filter(prompt: str, flags: dict[str, bool]) -> str:
+        """Keep unflagged lines; keep flagged lines only when a matching flag is set."""
+        lines = []
+        for line in prompt.split("\n"):
+            active_flags = _FLAG_PATTERN.findall(line)
+            cleaned = _FLAG_PATTERN.sub("", line).lstrip()
+            if not active_flags or any(flags.get(f, False) for f in active_flags):
+                lines.append(cleaned)
+        return "\n".join(lines)
+
+    @staticmethod
+    def _check_required_vars(prompt: str, formats: dict, prompt_name: str) -> None:
+        """Raise when any ``{var}`` placeholder lacks a corresponding kwarg."""
+        required = {f for _, f, _, _ in Formatter().parse(prompt) if f is not None}
+        if missing := required - set(formats.keys()):
+            raise ValueError(
+                f"Missing format variables for '{prompt_name}': {sorted(missing)}"
+            )
+
     def __repr__(self) -> str:
         return f"PromptHandler(language='{self.language}', num_prompts={len(self.data)})"
diff --git a/reme4/components/runtime_context.py b/reme4/components/runtime_context.py
index 8d8f4409..d9454900 100644
--- a/reme4/components/runtime_context.py
+++ b/reme4/components/runtime_context.py
@@ -9,8 +9,8 @@ from ..schema import Response, StreamChunk
 class RuntimeContext:
     """Scratch space for a single execution.
 
-    Holds the response object, an optional stream queue, and a free-form
-    data dict accessed via mapping-style operators.
+    Holds the response object, an optional stream queue, a stop event, and a
+    free-form data dict accessible via mapping-style operators (``ctx[key]``).
     """
 
     def __init__(
@@ -25,12 +25,14 @@ class RuntimeContext:
         self.stop_event: asyncio.Event | None = stop_event
         self.data: dict = kwargs
 
+    # ----- Data dict access ----------------------------------------------
+
     def get(self, key: str, default=None):
-        """Get a value from the data dict."""
+        """Get a value from the data dict with an optional default."""
         return self.data.get(key, default)
 
     def update(self, data: dict) -> "RuntimeContext":
-        """Merge data into the context."""
+        """Merge `data` into the context and return self for chaining."""
         self.data.update(data)
         return self
 
@@ -46,41 +48,40 @@ class RuntimeContext:
     def __contains__(self, key: str) -> bool:
         return key in self.data
 
-    @property
-    def stream(self) -> bool:
-        """Whether streaming is enabled."""
-        return self.stream_queue is not None
-
     @classmethod
     def from_context(cls, context: "RuntimeContext | None" = None, **kwargs) -> "RuntimeContext":
-        """Reuse or create a RuntimeContext."""
-        # Reuse the existing context (merging kwargs) or create a new one.
+        """Reuse `context` (merging kwargs into its data) or create a fresh one."""
         if context is None:
             return cls(**kwargs)
-        context.update(kwargs)
-        return context
+        return context.update(kwargs)
+
+    # ----- Streaming -----------------------------------------------------
+
+    @property
+    def stream(self) -> bool:
+        """Whether a stream queue is attached (i.e., streaming is enabled)."""
+        return self.stream_queue is not None
 
     async def _enqueue(self, chunk: StreamChunk) -> None:
-        """Put a chunk on the stream queue."""
+        """Put a chunk on the stream queue; raises when streaming is disabled."""
         if self.stream_queue is None:
             raise RuntimeError("Stream queue not initialized")
         await self.stream_queue.put(chunk)
 
     async def add_stream_string(self, chunk: str, chunk_type: ChunkEnum) -> "RuntimeContext":
         """Emit a text chunk to the stream queue."""
-        # Emit a text chunk to the stream queue.
         await self._enqueue(StreamChunk(chunk_type=chunk_type, chunk=chunk))
         return self
 
     async def add_stream_done(self) -> "RuntimeContext":
         """Emit the terminal DONE marker to close the stream."""
-        # Emit the terminal DONE marker to close the stream.
         await self._enqueue(StreamChunk(chunk_type=ChunkEnum.DONE, chunk="", done=True))
         return self
 
+    # ----- Misc ----------------------------------------------------------
+
     def apply_mapping(self, mapping: dict[str, str]) -> "RuntimeContext":
-        """Copy data[source] into data[target] for each mapping pair."""
-        # Copy data[source] into data[target] for each {source: target} pair.
+        """Copy ``data[source]`` to ``data[target]`` for each ``source: target`` pair."""
         if not mapping:
             return self
         for source, target in mapping.items():
diff --git a/reme4/components/service/base_service.py b/reme4/components/service/base_service.py
index 8d4d2402..2384fe71 100644
--- a/reme4/components/service/base_service.py
+++ b/reme4/components/service/base_service.py
@@ -60,9 +60,9 @@ class BaseService(BaseComponent):
         return lifespan
 
     def add_jobs(self, app: "Application") -> None:
-        """Register every job from the app context except background-only ones."""
+        """Register every job whose enable_serve flag is True."""
         for name, job in app.context.jobs.items():
-            if job.backend == "background":
+            if not job.enable_serve:
                 continue
             try:
                 self.add_job(job)
diff --git a/reme4/components/tokenizer/jieba_tokenizer.py b/reme4/components/tokenizer/jieba_tokenizer.py
index 1dc98b13..e63c7a71 100644
--- a/reme4/components/tokenizer/jieba_tokenizer.py
+++ b/reme4/components/tokenizer/jieba_tokenizer.py
@@ -1,15 +1,43 @@
 """Jieba tokenizer for Chinese text segmentation."""
 
+from typing import Callable
+
 from .base_tokenizer import BaseTokenizer
 from ..component_registry import R
 
 
 @R.register("jieba")
 class JiebaTokenizer(BaseTokenizer):
-    """Tokenizer backed by jieba for Chinese word segmentation."""
+    """Tokenizer backed by jieba for Chinese word segmentation.
+
+    `backend` selects the underlying implementation:
+      - "rjieba": Rust binding of jieba-rs, ~10-30x faster than pure Python (default).
+      - "jieba":  Original pure-Python jieba, slowest but the reference.
+    """
+
+    SUPPORTED_BACKENDS = ("rjieba", "jieba")
+
+    def __init__(self, backend: str = "rjieba", **kwargs):
+        super().__init__(**kwargs)
+        if backend not in self.SUPPORTED_BACKENDS:
+            raise ValueError(
+                f"Unknown jieba backend {backend!r}; expected one of {self.SUPPORTED_BACKENDS}"
+            )
+        self.backend = backend
+        self._cut: Callable[[str], list[str]] | None = None
+
+    async def _start(self) -> None:
+        await super()._start()
+        # Resolve the backend once at startup so per-call overhead is just one attribute lookup.
+        if self.backend == "rjieba":
+            import rjieba
+
+            self._cut = rjieba.cut
+        else:
+            import jieba
+
+            self._cut = jieba.cut
+        self.logger.info(f"JiebaTokenizer using backend: {self.backend}")
 
     def _tokenize_one(self, text: str, **kwargs) -> list[str]:
-        # Lazy import: jieba startup cost is non-trivial and only paid when used.
-        import jieba
-
-        return list(jieba.cut(text))
+        return list(self._cut(text))
diff --git a/reme4/config/default.yaml b/reme4/config/default.yaml
index 73afdbe8..95bb8369 100644
--- a/reme4/config/default.yaml
+++ b/reme4/config/default.yaml
@@ -14,8 +14,8 @@ jobs:
   # ════════════════════════════════════════════════════════════════════
   # UTILITY — service introspection
   # ════════════════════════════════════════════════════════════════════
-  - backend: base
-    name: version
+  version:
+    backend: base
     description: "return reme4 package version"
     parameters:
       type: object
@@ -23,8 +23,8 @@ jobs:
     steps:
       - backend: version_step
 
-  - backend: base
-    name: health_check
+  health_check:
+    backend: base
     description: "return a concise health-check snapshot of reme4 components"
     parameters:
       type: object
@@ -32,8 +32,8 @@ jobs:
     steps:
       - backend: health_check_step
 
-  - backend: base
-    name: help
+  help:
+    backend: base
     description: "list all registered jobs with their metadata"
     parameters:
       type: object
@@ -41,8 +41,8 @@ jobs:
     steps:
       - backend: help_step
 
-  - backend: base
-    name: reindex
+  reindex:
+    backend: base
     description: "wipe the file store and rebuild it from the watcher's tracked files"
     parameters:
       type: object
@@ -50,8 +50,8 @@ jobs:
     steps:
       - backend: reindex_step
 
-  - backend: base
-    name: index_changes
+  index_changes:
+    backend: base
     description: "apply a batch of file changes (added/modified/deleted) into file_store"
     parameters:
       type: object
@@ -82,8 +82,8 @@ jobs:
   # ════════════════════════════════════════════════════════════════════
 
   # ── Retrieve ───────────────────────────────────────────────────────
-  - backend: base
-    name: search
+  search:
+    backend: base
     description: "Hybrid vault search (vector + BM25, RRF-fused)."
     parameters:
       type: object
@@ -108,8 +108,8 @@ jobs:
         expand_links: true
         max_links_per_direction: 10
 
-  - backend: base
-    name: traverse
+  traverse:
+    backend: base
     description: "Walk the wikilink graph from a seed path."
     parameters:
       type: object
@@ -131,8 +131,8 @@ jobs:
       - backend: traverse_step
 
   # ── Read Operations ───────────────────────────────────────────────────────────
-  - backend: base
-    name: list
+  list:
+    backend: base
     description: "List files under a vault path."
     parameters:
       type: object
@@ -152,8 +152,8 @@ jobs:
     steps:
       - backend: list_step
 
-  - backend: base
-    name: read
+  read:
+    backend: base
     description: "Read a markdown file under the vault."
     parameters:
       type: object
@@ -172,8 +172,8 @@ jobs:
     steps:
       - backend: read_step
 
-  - backend: base
-    name: stat
+  stat:
+    backend: base
     description: "Stat a vault file (size, mtime, exists, is_dir, is_file)."
     parameters:
       type: object
@@ -186,8 +186,8 @@ jobs:
     steps:
       - backend: stat_step
 
-  - backend: base
-    name: frontmatter:read
+  frontmatter:read:
+    backend: base
     description: "Read a file's YAML frontmatter as a dict."
     parameters:
       type: object
@@ -201,8 +201,8 @@ jobs:
       - backend: frontmatter:read_step
 
   # ── Write Operations──────────────────────────────────────────────────────────
-  - backend: base
-    name: write
+  write:
+    backend: base
     description: "Write a markdown file (create or overwrite) with name/description frontmatter."
     parameters:
       type: object
@@ -227,8 +227,8 @@ jobs:
     steps:
       - backend: write_step
 
-  - backend: base
-    name: edit
+  edit:
+    backend: base
     description: "Find-and-replace in a markdown file (all occurrences)."
     parameters:
       type: object
@@ -250,8 +250,8 @@ jobs:
     steps:
       - backend: edit_step
 
-  - backend: base
-    name: append
+  append:
+    backend: base
     description: "Append content to a markdown file."
     parameters:
       type: object
@@ -268,8 +268,8 @@ jobs:
     steps:
       - backend: append_step
 
-  - backend: base
-    name: frontmatter:update
+  frontmatter:update:
+    backend: base
     description: "Merge keys into a file's YAML frontmatter."
     parameters:
       type: object
@@ -287,8 +287,8 @@ jobs:
     steps:
       - backend: frontmatter_update_step
 
-  - backend: base
-    name: frontmatter:delete
+  frontmatter:delete:
+    backend: base
     description: "Drop keys from a file's YAML frontmatter."
     parameters:
       type: object
@@ -308,8 +308,8 @@ jobs:
       - backend: frontmatter_delete_step
 
   # ── File Operations (relocate / cross-realm) ──────────────────────────────
-  - backend: base
-    name: move
+  move:
+    backend: base
     description: "Move / rename a vault file; rewrites inbound wikilinks by default."
     parameters:
       type: object
@@ -334,8 +334,8 @@ jobs:
     steps:
       - backend: move_step
 
-  - backend: base
-    name: delete
+  delete:
+    backend: base
     description: "Delete a vault file or folder; returns surviving inbound wikilinks."
     parameters:
       type: object
@@ -348,8 +348,8 @@ jobs:
     steps:
       - backend: delete_step
 
-  - backend: base
-    name: upload
+  upload:
+    backend: base
     description: "Copy a host file into the vault at an explicit destination."
     parameters:
       type: object
@@ -370,8 +370,8 @@ jobs:
     steps:
       - backend: upload_step
 
-  - backend: base
-    name: upload_resource
+  upload_resource:
+    backend: base
     description: "Ingest an external-channel asset into resource/<today>/ with provenance."
     parameters:
       type: object
@@ -396,8 +396,8 @@ jobs:
     steps:
       - backend: upload_resource_step
 
-  - backend: base
-    name: download
+  download:
+    backend: base
     description: "Copy a vault file out to the host filesystem."
     parameters:
       type: object
@@ -419,8 +419,8 @@ jobs:
       - backend: download_step
 
   # ── Daily Operations (note CRUD + day-index rollup) ───────────────────
-  - backend: base
-    name: daily:read
+  daily:read:
+    backend: base
     description: "Read daily/<date>/<slug>.md (body + frontmatter)."
     parameters:
       type: object
@@ -437,8 +437,8 @@ jobs:
     steps:
       - backend: daily_read_step
 
-  - backend: base
-    name: daily:write
+  daily:write:
+    backend: base
     description: "Write daily/<date>/<slug>.md (body + frontmatter); refreshes the day index."
     parameters:
       type: object
@@ -471,8 +471,8 @@ jobs:
     steps:
       - backend: daily_write_step
 
-  - backend: base
-    name: daily:list
+  daily:list:
+    backend: base
     description: "List notes under a single day."
     parameters:
       type: object
@@ -484,8 +484,8 @@ jobs:
     steps:
       - backend: daily_list_step
 
-  - backend: base
-    name: daily:reindex
+  daily:reindex:
+    backend: base
     description: "Rebuild the day-index page daily/<date>.md."
     parameters:
       type: object
@@ -497,8 +497,8 @@ jobs:
     steps:
       - backend: daily_reindex_step
 
-  - backend: background
-    name: watch_file
+  watch_file:
+    backend: background
     watch_paths:
       - MEMORY.md
       - memory
diff --git a/reme4/config/demo.yaml b/reme4/config/demo.yaml
index 6fce17eb..51e6bd5d 100644
--- a/reme4/config/demo.yaml
+++ b/reme4/config/demo.yaml
@@ -2,8 +2,8 @@ service:
   backend: http
 
 jobs:
-  - backend: base
-    name: version
+  version:
+    backend: base
     description: "return reme4 package version"
     parameters:
       type: object
@@ -11,8 +11,8 @@ jobs:
     steps:
       - backend: version_step
 
-  - backend: base
-    name: help
+  help:
+    backend: base
     description: "list all registered jobs with their metadata"
     parameters:
       type: object
@@ -20,8 +20,8 @@ jobs:
     steps:
       - backend: help_step
 
-  - backend: base
-    name: demo
+  demo:
+    backend: base
     description: "demo job description"
     parameters:
       type: object
@@ -39,8 +39,8 @@ jobs:
       - backend: demo_echo_step1
       - backend: demo_echo_step2
 
-  - backend: stream
-    name: stream_demo
+  stream_demo:
+    backend: stream
     description: "stream demo job: repeat query 10x and stream char-by-char"
     parameters:
       type: object
diff --git a/reme4/config/qwenpaw.yaml b/reme4/config/qwenpaw.yaml
index 73afdbe8..95bb8369 100644
--- a/reme4/config/qwenpaw.yaml
+++ b/reme4/config/qwenpaw.yaml
@@ -14,8 +14,8 @@ jobs:
   # ════════════════════════════════════════════════════════════════════
   # UTILITY — service introspection
   # ════════════════════════════════════════════════════════════════════
-  - backend: base
-    name: version
+  version:
+    backend: base
     description: "return reme4 package version"
     parameters:
       type: object
@@ -23,8 +23,8 @@ jobs:
     steps:
       - backend: version_step
 
-  - backend: base
-    name: health_check
+  health_check:
+    backend: base
     description: "return a concise health-check snapshot of reme4 components"
     parameters:
       type: object
@@ -32,8 +32,8 @@ jobs:
     steps:
       - backend: health_check_step
 
-  - backend: base
-    name: help
+  help:
+    backend: base
     description: "list all registered jobs with their metadata"
     parameters:
       type: object
@@ -41,8 +41,8 @@ jobs:
     steps:
       - backend: help_step
 
-  - backend: base
-    name: reindex
+  reindex:
+    backend: base
     description: "wipe the file store and rebuild it from the watcher's tracked files"
     parameters:
       type: object
@@ -50,8 +50,8 @@ jobs:
     steps:
       - backend: reindex_step
 
-  - backend: base
-    name: index_changes
+  index_changes:
+    backend: base
     description: "apply a batch of file changes (added/modified/deleted) into file_store"
     parameters:
       type: object
@@ -82,8 +82,8 @@ jobs:
   # ════════════════════════════════════════════════════════════════════
 
   # ── Retrieve ───────────────────────────────────────────────────────
-  - backend: base
-    name: search
+  search:
+    backend: base
     description: "Hybrid vault search (vector + BM25, RRF-fused)."
     parameters:
       type: object
@@ -108,8 +108,8 @@ jobs:
         expand_links: true
         max_links_per_direction: 10
 
-  - backend: base
-    name: traverse
+  traverse:
+    backend: base
     description: "Walk the wikilink graph from a seed path."
     parameters:
       type: object
@@ -131,8 +131,8 @@ jobs:
       - backend: traverse_step
 
   # ── Read Operations ───────────────────────────────────────────────────────────
-  - backend: base
-    name: list
+  list:
+    backend: base
     description: "List files under a vault path."
     parameters:
       type: object
@@ -152,8 +152,8 @@ jobs:
     steps:
       - backend: list_step
 
-  - backend: base
-    name: read
+  read:
+    backend: base
     description: "Read a markdown file under the vault."
     parameters:
       type: object
@@ -172,8 +172,8 @@ jobs:
     steps:
       - backend: read_step
 
-  - backend: base
-    name: stat
+  stat:
+    backend: base
     description: "Stat a vault file (size, mtime, exists, is_dir, is_file)."
     parameters:
       type: object
@@ -186,8 +186,8 @@ jobs:
     steps:
       - backend: stat_step
 
-  - backend: base
-    name: frontmatter:read
+  frontmatter:read:
+    backend: base
     description: "Read a file's YAML frontmatter as a dict."
     parameters:
       type: object
@@ -201,8 +201,8 @@ jobs:
       - backend: frontmatter:read_step
 
   # ── Write Operations──────────────────────────────────────────────────────────
-  - backend: base
-    name: write
+  write:
+    backend: base
     description: "Write a markdown file (create or overwrite) with name/description frontmatter."
     parameters:
       type: object
@@ -227,8 +227,8 @@ jobs:
     steps:
       - backend: write_step
 
-  - backend: base
-    name: edit
+  edit:
+    backend: base
     description: "Find-and-replace in a markdown file (all occurrences)."
     parameters:
       type: object
@@ -250,8 +250,8 @@ jobs:
     steps:
       - backend: edit_step
 
-  - backend: base
-    name: append
+  append:
+    backend: base
     description: "Append content to a markdown file."
     parameters:
       type: object
@@ -268,8 +268,8 @@ jobs:
     steps:
       - backend: append_step
 
-  - backend: base
-    name: frontmatter:update
+  frontmatter:update:
+    backend: base
     description: "Merge keys into a file's YAML frontmatter."
     parameters:
       type: object
@@ -287,8 +287,8 @@ jobs:
     steps:
       - backend: frontmatter_update_step
 
-  - backend: base
-    name: frontmatter:delete
+  frontmatter:delete:
+    backend: base
     description: "Drop keys from a file's YAML frontmatter."
     parameters:
       type: object
@@ -308,8 +308,8 @@ jobs:
       - backend: frontmatter_delete_step
 
   # ── File Operations (relocate / cross-realm) ──────────────────────────────
-  - backend: base
-    name: move
+  move:
+    backend: base
     description: "Move / rename a vault file; rewrites inbound wikilinks by default."
     parameters:
       type: object
@@ -334,8 +334,8 @@ jobs:
     steps:
       - backend: move_step
 
-  - backend: base
-    name: delete
+  delete:
+    backend: base
     description: "Delete a vault file or folder; returns surviving inbound wikilinks."
     parameters:
       type: object
@@ -348,8 +348,8 @@ jobs:
     steps:
       - backend: delete_step
 
-  - backend: base
-    name: upload
+  upload:
+    backend: base
     description: "Copy a host file into the vault at an explicit destination."
     parameters:
       type: object
@@ -370,8 +370,8 @@ jobs:
     steps:
       - backend: upload_step
 
-  - backend: base
-    name: upload_resource
+  upload_resource:
+    backend: base
     description: "Ingest an external-channel asset into resource/<today>/ with provenance."
     parameters:
       type: object
@@ -396,8 +396,8 @@ jobs:
     steps:
       - backend: upload_resource_step
 
-  - backend: base
-    name: download
+  download:
+    backend: base
     description: "Copy a vault file out to the host filesystem."
     parameters:
       type: object
@@ -419,8 +419,8 @@ jobs:
       - backend: download_step
 
   # ── Daily Operations (note CRUD + day-index rollup) ───────────────────
-  - backend: base
-    name: daily:read
+  daily:read:
+    backend: base
     description: "Read daily/<date>/<slug>.md (body + frontmatter)."
     parameters:
       type: object
@@ -437,8 +437,8 @@ jobs:
     steps:
       - backend: daily_read_step
 
-  - backend: base
-    name: daily:write
+  daily:write:
+    backend: base
     description: "Write daily/<date>/<slug>.md (body + frontmatter); refreshes the day index."
     parameters:
       type: object
@@ -471,8 +471,8 @@ jobs:
     steps:
       - backend: daily_write_step
 
-  - backend: base
-    name: daily:list
+  daily:list:
+    backend: base
     description: "List notes under a single day."
     parameters:
       type: object
@@ -484,8 +484,8 @@ jobs:
     steps:
       - backend: daily_list_step
 
-  - backend: base
-    name: daily:reindex
+  daily:reindex:
+    backend: base
     description: "Rebuild the day-index page daily/<date>.md."
     parameters:
       type: object
@@ -497,8 +497,8 @@ jobs:
     steps:
       - backend: daily_reindex_step
 
-  - backend: background
-    name: watch_file
+  watch_file:
+    backend: background
     watch_paths:
       - MEMORY.md
       - memory
diff --git a/reme4/schema/application_config.py b/reme4/schema/application_config.py
index 6e82e527..3df6ff1d 100644
--- a/reme4/schema/application_config.py
+++ b/reme4/schema/application_config.py
@@ -16,12 +16,12 @@ class ComponentConfig(BaseModel):
 
 
 class JobConfig(ComponentConfig):
-    """Config for a job — an ordered sequence of step components."""
+    """Config for a job — an ordered sequence of step components. Keyed by name in ApplicationConfig.jobs."""
 
-    name: str = Field(default="", description="Unique job identifier")
     description: str = Field(default="", description="Human-readable description")
     parameters: dict = Field(default_factory=dict, description="Job-level parameters")
     steps: list[ComponentConfig] = Field(default_factory=list, description="Ordered step configs")
+    enable_serve: bool = Field(default=True, description="Whether to expose this job through the service layer")
 
 
 class ApplicationConfig(BaseModel):
@@ -42,7 +42,10 @@ class ApplicationConfig(BaseModel):
     log_to_file: bool = Field(default=True, description="Log to file")
     mcp_servers: dict[str, dict] = Field(default_factory=dict, description="MCP server configs by name")
     service: ComponentConfig = Field(default_factory=ComponentConfig, description="Service endpoint config")
-    jobs: list[JobConfig] = Field(default_factory=list, description="Job definitions")
+    jobs: dict[str, JobConfig] = Field(
+        default_factory=dict,
+        description="Job definitions keyed by job name",
+    )
     components: dict[ComponentEnum, dict[str, ComponentConfig]] = Field(
         default_factory=dict,
         description="Component registry keyed by type then name",