Merge branch 'master' of http://192.168.0.200/knowledge-base-bj/deerflow-server into dhk-1

offline-backend-20260512/backend/app/gateway/app.py

@@ -53,6 +53,7 @@ from app.gateway.routers import (
     thread_shares,
     threads,
     uploads,
+    user_prompts,
     writing,
 )
 from deerflow.config import app_config as deerflow_app_config
@@ -824,6 +825,8 @@ This gateway provides custom endpoints for models, MCP configuration, skills, an
     # Recommended Questions API is mounted at /api/recommended-questions
     app.include_router(recommended_questions.router)
 
+    app.include_router(user_prompts.router)
+
     # Suggestions API is mounted at /api/threads/{thread_id}/suggestions
     app.include_router(suggestions.router)
 

offline-backend-20260512/backend/app/gateway/deps.py

@@ -148,6 +148,10 @@ async def langgraph_runtime(app: FastAPI) -> AsyncGenerator[None, None]:
 
         app.state.memory_config_store = make_user_memory_config_store(sf)
 
+        from deerflow.persistence.user_prompts import make_user_prompt_store
+
+        app.state.user_prompt_store = make_user_prompt_store(sf)
+
         # Run event store (has its own factory with config-driven backend selection)
         run_events_config = getattr(config, "run_events", None)
         app.state.run_event_store = make_run_event_store(run_events_config)

offline-backend-20260512/backend/app/gateway/routers/agents.py

@@ -59,6 +59,11 @@ class AgentResponse(BaseModel):
         description="Admin-curated homepage rank. Null = not featured. Lower sorts first.",
     )
 
+    is_favorite: bool = Field(
+        default=False,
+        description="Whether the requesting user has favorited (收藏) this agent.",
+    )
+
 
 class AgentsListResponse(BaseModel):
     """Response model for listing all visible agents."""
@@ -440,8 +445,10 @@ async def list_agents(
         ]
 
         assignments = await tag_store.list_assignments("agent", [r.id for r in responses])
+        favorite_ids = await store.list_favorite_ids(user_id)
         for resp in responses:
             resp.tags = assignments.get(resp.id, [])
+            resp.is_favorite = resp.id in favorite_ids
         if search:
             needle = search.strip().lower()
             responses = [r for r in responses if needle in r.name.lower()]
@@ -520,7 +527,10 @@ async def get_agent(agent_id: str, request: Request) -> AgentResponse:
         # Detail view legitimately reads SOUL.md + config.yaml — hop to a
         # worker thread so the asyncio loop stays responsive under concurrent
         # detail fetches.
-        return await _record_to_response_async(record, include_soul=True, include_extras=True)
+        response = await _record_to_response_async(record, include_soul=True, include_extras=True)
+        favorite_ids = await store.list_favorite_ids(user_id)
+        response.is_favorite = agent_id in favorite_ids
+        return response
     except HTTPException:
         raise
     except Exception as e:
@@ -747,6 +757,10 @@ async def delete_agent(agent_id: str, request: Request) -> None:
             await get_tag_store(request).unassign_all("agent", agent_id)
         except Exception:  # noqa: BLE001 — tag cleanup must not block deletion
             pass
+        try:
+            await store.delete_favorites_for_agent(agent_id)
+        except Exception:  # noqa: BLE001 — favorite cleanup must not block deletion
+            pass
         logger.info("Deleted agent '%s' from %s", agent_id, agent_dir)
     except HTTPException:
         raise
@@ -755,6 +769,53 @@ async def delete_agent(agent_id: str, request: Request) -> None:
         raise HTTPException(status_code=500, detail=f"Failed to delete agent: {str(e)}") from e
 
 
+@router.post(
+    "/agents/{agent_id}/favorite",
+    status_code=204,
+    summary="Favorite Agent",
+    description="Bookmark a visible agent into the current user's 我的 tab. Idempotent.",
+)
+async def favorite_agent(agent_id: str, request: Request) -> None:
+    _require_agents_api_enabled()
+    agent_id = _validate_storage_id(agent_id)
+
+    try:
+        store = get_agent_store(request)
+        user_id = _current_user_id(request)
+        record = await store.get_visible(agent_id, user_id)
+        if record is None:
+            raise HTTPException(status_code=404, detail=f"Agent '{agent_id}' not found")
+        await store.add_favorite(user_id, agent_id)
+        logger.info("User '%s' favorited agent '%s'", user_id, agent_id)
+    except HTTPException:
+        raise
+    except Exception as e:
+        logger.error("Failed to favorite agent '%s': %s", agent_id, e, exc_info=True)
+        raise HTTPException(status_code=500, detail=f"Failed to favorite agent: {str(e)}") from e
+
+
+@router.delete(
+    "/agents/{agent_id}/favorite",
+    status_code=204,
+    summary="Unfavorite Agent",
+    description="Remove an agent from the current user's favorites. No-op when not favorited.",
+)
+async def unfavorite_agent(agent_id: str, request: Request) -> None:
+    _require_agents_api_enabled()
+    agent_id = _validate_storage_id(agent_id)
+
+    try:
+        store = get_agent_store(request)
+        user_id = _current_user_id(request)
+        await store.remove_favorite(user_id, agent_id)
+        logger.info("User '%s' unfavorited agent '%s'", user_id, agent_id)
+    except HTTPException:
+        raise
+    except Exception as e:
+        logger.error("Failed to unfavorite agent '%s': %s", agent_id, e, exc_info=True)
+        raise HTTPException(status_code=500, detail=f"Failed to unfavorite agent: {str(e)}") from e
+
+
 @router.put(
     "/agents/{agent_id}/featured",
     response_model=AgentResponse,

offline-backend-20260512/backend/app/gateway/routers/scheduled_tasks.py

@@ -37,6 +37,10 @@ class ScheduledTaskCreateRequest(BaseModel):
     timezone: str | None = None
     execution_agent_name: str | None = None
     execution_context: dict[str, Any] | None = None
+    require_markdown: bool = Field(
+        default=False,
+        description="Require the result to be Markdown and exceed the minimum length, else mark the run failed.",
+    )
 
 
 class ScheduledTaskUpdateRequest(BaseModel):
@@ -48,6 +52,7 @@ class ScheduledTaskUpdateRequest(BaseModel):
     enabled: bool | None = None
     execution_agent_name: str | None = None
     execution_context: dict[str, Any] | None = None
+    require_markdown: bool | None = None
 
 
 class ScheduledTaskDeliveryProfileUpdateRequest(BaseModel):
@@ -555,6 +560,7 @@ async def create_scheduled_task(body: ScheduledTaskCreateRequest, request: Reque
             timezone=body.timezone,
             execution_agent_name=body.execution_agent_name,
             execution_context=body.execution_context,
+            require_markdown=body.require_markdown,
         )
     except DuplicateTaskNameError:
         raise HTTPException(status_code=409, detail="已存在同名定时任务,请换一个名称")

offline-backend-20260512/backend/app/gateway/routers/threads.py

@@ -35,6 +35,7 @@ from deerflow.runtime import serialize_channel_values
 from deerflow.runtime.references import (
     adapt_reference_batches_with_displays,
     build_reference_batches as build_runtime_reference_batches,
+    build_reference_debug_payload,
     build_reference_debug_summary,
 )
 from deerflow.runtime.user_context import get_effective_user_id
@@ -231,6 +232,7 @@ class ReferenceSourceResponse(BaseModel):
     type: str = "result"
     snippet: str = ""
     source: str | None = None
+    author: str | None = None
     time: str | None = None
     url: str | None = None
     content: str | None = None
@@ -843,7 +845,8 @@ async def get_thread_reference_batches(thread_id: str, request: Request) -> Refe
             _rag_debug_file_path(),
         )
         return ReferenceBatchesResponse(batches=[])
-    raw_batches = build_runtime_reference_batches(messages)
+    batches_before_display_adapt = build_runtime_reference_batches(messages)
+    raw_batches = batches_before_display_adapt
     persisted_displays = await _load_persisted_skill_displays()
     if persisted_displays:
         raw_batches = adapt_reference_batches_with_displays(raw_batches, persisted_displays)
@@ -866,6 +869,25 @@ async def get_thread_reference_batches(thread_id: str, request: Request) -> Refe
         json.dumps(debug_summary["turns_detail"], ensure_ascii=False),
         _rag_debug_file_path(),
     )
+    _log_rag_debug(
+        "backend reference-batches-payload thread=%s %s",
+        sanitize_log_param(thread_id),
+        json.dumps(
+            {
+                "display_adapter_names": sorted(persisted_displays.keys()),
+                "before_display_adapt": build_reference_debug_payload(
+                    messages=messages,
+                    batches=batches_before_display_adapt,
+                ),
+                "after_display_adapt": build_reference_debug_payload(
+                    messages=messages,
+                    batches=raw_batches,
+                ),
+            },
+            ensure_ascii=False,
+            default=str,
+        ),
+    )
     return ReferenceBatchesResponse(
         batches=[
             ReferenceBatchResponse(

offline-backend-20260512/backend/app/gateway/routers/user_prompts.py

@@ -0,0 +1,108 @@
+"""CRUD API for a user's personal custom prompts (常用提示词).
+
+Prompts are bound to the requesting user (not stored in the browser), so they
+follow the account across devices/sessions.
+"""
+
+from __future__ import annotations
+
+import logging
+from datetime import datetime
+from uuid import uuid4
+
+from fastapi import APIRouter, HTTPException, Request
+from pydantic import BaseModel, Field
+
+from deerflow.runtime.user_context import get_effective_user_id
+
+logger = logging.getLogger(__name__)
+router = APIRouter(prefix="/api/user-prompts", tags=["user-prompts"])
+
+
+class UserPromptResponse(BaseModel):
+    id: str
+    title: str = ""
+    content: str
+    sort_order: int = 0
+    created_at: datetime | str | None = None
+    updated_at: datetime | str | None = None
+
+
+class UserPromptsListResponse(BaseModel):
+    prompts: list[UserPromptResponse]
+
+
+class UserPromptCreateRequest(BaseModel):
+    title: str = Field(default="", max_length=255)
+    content: str = Field(..., min_length=1)
+    sort_order: int = Field(default=0)
+
+
+class UserPromptUpdateRequest(BaseModel):
+    title: str | None = Field(default=None, max_length=255)
+    content: str | None = Field(default=None, min_length=1)
+    sort_order: int | None = None
+
+
+def _current_user_id(request: Request) -> str:
+    user = getattr(request.state, "user", None)
+    if user is not None:
+        return str(user.id)
+    return get_effective_user_id()
+
+
+def _get_store(request: Request):
+    store = getattr(request.app.state, "user_prompt_store", None)
+    if store is None:
+        raise HTTPException(status_code=503, detail="User prompt store not available")
+    return store
+
+
+@router.get("", response_model=UserPromptsListResponse)
+async def list_user_prompts(request: Request) -> UserPromptsListResponse:
+    store = _get_store(request)
+    user_id = _current_user_id(request)
+    rows = await store.list_for_user(user_id)
+    return UserPromptsListResponse(prompts=[UserPromptResponse(**r) for r in rows])
+
+
+@router.post("", response_model=UserPromptResponse, status_code=201)
+async def create_user_prompt(request: Request, body: UserPromptCreateRequest) -> UserPromptResponse:
+    store = _get_store(request)
+    user_id = _current_user_id(request)
+    row = await store.create(
+        user_id,
+        {
+            "id": uuid4().hex,
+            "title": (body.title or "").strip(),
+            "content": body.content,
+            "sort_order": body.sort_order,
+        },
+    )
+    return UserPromptResponse(**row)
+
+
+@router.put("/{prompt_id}", response_model=UserPromptResponse)
+async def update_user_prompt(request: Request, prompt_id: str, body: UserPromptUpdateRequest) -> UserPromptResponse:
+    store = _get_store(request)
+    user_id = _current_user_id(request)
+    data: dict = {}
+    if body.title is not None:
+        data["title"] = body.title.strip()
+    if body.content is not None:
+        data["content"] = body.content
+    if body.sort_order is not None:
+        data["sort_order"] = body.sort_order
+    row = await store.update(prompt_id, user_id, data)
+    if row is None:
+        raise HTTPException(status_code=404, detail="Prompt not found")
+    return UserPromptResponse(**row)
+
+
+@router.delete("/{prompt_id}", status_code=204)
+async def delete_user_prompt(request: Request, prompt_id: str) -> None:
+    store = _get_store(request)
+    user_id = _current_user_id(request)
+    deleted = await store.delete(prompt_id, user_id)
+    if not deleted:
+        raise HTTPException(status_code=404, detail="Prompt not found")

offline-backend-20260512/backend/packages/harness/deerflow/agents/middlewares/rag_citation_middleware.py

@@ -2,6 +2,7 @@
 
 from __future__ import annotations
 
+import json
 import logging
 import os
 from pathlib import Path
@@ -13,7 +14,11 @@ from langchain.agents.middleware import AgentMiddleware
 from langchain.agents.middleware.types import ModelCallResult, ModelRequest, ModelResponse
 from langchain_core.messages import SystemMessage
 
-from deerflow.runtime.references import build_latest_reference_prompt, build_reference_debug_summary
+from deerflow.runtime.references import (
+    build_latest_reference_prompt,
+    build_reference_debug_payload,
+    build_reference_debug_summary,
+)
 
 logger = logging.getLogger(__name__)
 _RAG_DEBUG_LOGGER_NAME = "rag_citation_debug"
@@ -84,6 +89,18 @@ class RagCitationMiddleware(AgentMiddleware[AgentState]):
             summary["kept_items"],
             len((batch or {}).get("sources") or []),
         )
+        _log_rag_debug(
+            "middleware reference-prompt-payload %s",
+            json.dumps(
+                build_reference_debug_payload(
+                    messages=messages,
+                    batch=batch,
+                    prompt=prompt,
+                ),
+                ensure_ascii=False,
+                default=str,
+            ),
+        )
         patched_messages = [
             *messages,
             SystemMessage(content=prompt),

offline-backend-20260512/backend/packages/harness/deerflow/config/extensions_config.py

@@ -75,6 +75,7 @@ class SkillDisplayConfig(BaseModel):
     summary_field: str | None = Field(default=None, description="Field used as short summary/snippet", alias="summaryField")
     content_field: str | None = Field(default=None, description="Field used as full item content", alias="contentField")
     source_field: str | None = Field(default=None, description="Field used as source label", alias="sourceField")
+    author_field: str | None = Field(default=None, description="Field used as author/byline label", alias="authorField")
     type_field: str | None = Field(default=None, description="Field used as source/type badge", alias="typeField")
     time_field: str | None = Field(default=None, description="Field used as publication/update time", alias="timeField")
     score_field: str | None = Field(default=None, description="Field used as relevance score", alias="scoreField")

offline-backend-20260512/backend/packages/harness/deerflow/persistence/agents/__init__.py

@@ -4,10 +4,11 @@ from __future__ import annotations
 
 from deerflow.persistence.agents.base import AgentStore
 from deerflow.persistence.agents.memory import MemoryAgentStore
-from deerflow.persistence.agents.model import AgentRow
+from deerflow.persistence.agents.model import AgentFavoriteRow, AgentRow
 from deerflow.persistence.agents.sql import AgentRepository
 
 __all__ = [
+    "AgentFavoriteRow",
     "AgentRepository",
     "AgentRow",
     "AgentStore",

offline-backend-20260512/backend/packages/harness/deerflow/persistence/agents/base.py

@@ -117,3 +117,25 @@ class AgentStore(ABC):
         ``False`` if the agent does not exist.
         """
         raise NotImplementedError
+
+    # --- Favorites ("收藏") ---
+
+    @abstractmethod
+    async def add_favorite(self, user_id: str, agent_id: str) -> None:
+        """Bookmark ``agent_id`` for ``user_id``. Idempotent."""
+        raise NotImplementedError
+
+    @abstractmethod
+    async def remove_favorite(self, user_id: str, agent_id: str) -> bool:
+        """Remove a bookmark. Returns ``False`` when none existed."""
+        raise NotImplementedError
+
+    @abstractmethod
+    async def list_favorite_ids(self, user_id: str) -> set[str]:
+        """Return the set of agent ids ``user_id`` has favorited."""
+        raise NotImplementedError
+
+    @abstractmethod
+    async def delete_favorites_for_agent(self, agent_id: str) -> None:
+        """Drop every user's favorite of ``agent_id`` (called on agent delete)."""
+        raise NotImplementedError

offline-backend-20260512/backend/packages/harness/deerflow/persistence/agents/memory.py

@@ -21,6 +21,7 @@ def _copy(data: dict[str, Any]) -> dict[str, Any]:
 class MemoryAgentStore(AgentStore):
     def __init__(self) -> None:
         self._agents: dict[str, dict[str, Any]] = {}
+        self._favorites: set[tuple[str, str]] = set()
 
     @staticmethod
     def _is_visible(agent: dict[str, Any], user_id: str) -> bool:
@@ -164,3 +165,20 @@ class MemoryAgentStore(AgentStore):
             return False
         self._agents.pop(agent_id, None)
         return True
+
+    # --- Favorites ("收藏") ---
+
+    async def add_favorite(self, user_id: str, agent_id: str) -> None:
+        self._favorites.add((user_id, agent_id))
+
+    async def remove_favorite(self, user_id: str, agent_id: str) -> bool:
+        if (user_id, agent_id) in self._favorites:
+            self._favorites.discard((user_id, agent_id))
+            return True
+        return False
+
+    async def list_favorite_ids(self, user_id: str) -> set[str]:
+        return {agent_id for (uid, agent_id) in self._favorites if uid == user_id}
+
+    async def delete_favorites_for_agent(self, agent_id: str) -> None:
+        self._favorites = {(uid, aid) for (uid, aid) in self._favorites if aid != agent_id}

offline-backend-20260512/backend/packages/harness/deerflow/persistence/agents/model.py

@@ -4,7 +4,7 @@ from __future__ import annotations
 
 from datetime import UTC, datetime
 
-from sqlalchemy import BigInteger, Boolean, String
+from sqlalchemy import BigInteger, Boolean, String, UniqueConstraint
 from sqlalchemy.orm import Mapped, mapped_column
 
 from deerflow.persistence.base import Base
@@ -35,3 +35,20 @@ class AgentRow(Base):
     featured_order: Mapped[int | None] = mapped_column(BigInteger, nullable=True, index=True)
     created_at: Mapped[datetime] = mapped_column(BeijingDateTime(), nullable=False, default=lambda: datetime.now(UTC))
     updated_at: Mapped[datetime] = mapped_column(BeijingDateTime(), nullable=False, default=lambda: datetime.now(UTC), onupdate=lambda: datetime.now(UTC))
+
+
+class AgentFavoriteRow(Base):
+    """A user's bookmark of a (typically public-square) agent.
+
+    Lets a user collect agents from the 广场 (square) into their "我的" tab
+    without owning them. One row per (user, agent); the unique constraint makes
+    favoriting idempotent.
+    """
+
+    __tablename__ = "agent_favorites"
+    __table_args__ = (UniqueConstraint("user_id", "agent_id", name="uq_agent_favorites_user_agent"),)
+
+    id: Mapped[str] = mapped_column(String(64), primary_key=True)
+    user_id: Mapped[str] = mapped_column(String(64), nullable=False, index=True)
+    agent_id: Mapped[str] = mapped_column(String(64), nullable=False, index=True)
+    created_at: Mapped[datetime] = mapped_column(BeijingDateTime(), nullable=False, default=lambda: datetime.now(UTC))

offline-backend-20260512/backend/packages/harness/deerflow/persistence/agents/sql.py

@@ -4,12 +4,14 @@ from __future__ import annotations
 
 from datetime import UTC, datetime
 from typing import Any
+from uuid import uuid4
 
-from sqlalchemy import and_, func, or_, select
+from sqlalchemy import and_, delete, func, or_, select
+from sqlalchemy.exc import IntegrityError
 from sqlalchemy.ext.asyncio import AsyncSession, async_sessionmaker
 
 from deerflow.persistence.agents.base import AgentStore
-from deerflow.persistence.agents.model import AgentRow
+from deerflow.persistence.agents.model import AgentFavoriteRow, AgentRow
 
 
 def _row_to_dict(row: AgentRow) -> dict[str, Any]:
@@ -230,3 +232,47 @@ class AgentRepository(AgentStore):
             await session.delete(row)
             await session.commit()
             return True
+
+    # --- Favorites ("收藏") ---
+
+    async def add_favorite(self, user_id: str, agent_id: str) -> None:
+        async with self._sf() as session:
+            existing = (
+                await session.execute(
+                    select(AgentFavoriteRow.id)
+                    .where(AgentFavoriteRow.user_id == user_id, AgentFavoriteRow.agent_id == agent_id)
+                    .limit(1)
+                )
+            ).scalar_one_or_none()
+            if existing is not None:
+                return
+            session.add(AgentFavoriteRow(id=uuid4().hex, user_id=user_id, agent_id=agent_id))
+            try:
+                await session.commit()
+            except IntegrityError:
+                # Lost a race against a concurrent favorite of the same agent —
+                # the unique constraint already guarantees the desired state.
+                await session.rollback()
+
+    async def remove_favorite(self, user_id: str, agent_id: str) -> bool:
+        async with self._sf() as session:
+            result = await session.execute(
+                delete(AgentFavoriteRow).where(
+                    AgentFavoriteRow.user_id == user_id,
+                    AgentFavoriteRow.agent_id == agent_id,
+                )
+            )
+            await session.commit()
+            return (result.rowcount or 0) > 0
+
+    async def list_favorite_ids(self, user_id: str) -> set[str]:
+        async with self._sf() as session:
+            result = await session.execute(
+                select(AgentFavoriteRow.agent_id).where(AgentFavoriteRow.user_id == user_id)
+            )
+            return {row[0] for row in result.all()}
+
+    async def delete_favorites_for_agent(self, agent_id: str) -> None:
+        async with self._sf() as session:
+            await session.execute(delete(AgentFavoriteRow).where(AgentFavoriteRow.agent_id == agent_id))
+            await session.commit()

offline-backend-20260512/backend/packages/harness/deerflow/persistence/migrations/versions/20260528_01_agent_favorites.py

@@ -0,0 +1,45 @@
+"""Add ``agent_favorites`` for the agent collection (收藏) feature.
+
+Lets a user bookmark agents from the public square (广场) so they surface in
+their "我的" tab without owning them. One row per (user, agent).
+
+Revision ID: 20260528_01
+Revises: 20260526_01
+Create Date: 2026-05-28
+"""
+
+from __future__ import annotations
+
+import sqlalchemy as sa
+from alembic import op
+from sqlalchemy import inspect
+
+
+revision = "20260528_01"
+down_revision = "20260526_01"
+branch_labels = None
+depends_on = None
+
+
+def _has_table(table_name: str) -> bool:
+    inspector = inspect(op.get_bind())
+    return table_name in inspector.get_table_names()
+
+
+def upgrade() -> None:
+    if not _has_table("agent_favorites"):
+        op.create_table(
+            "agent_favorites",
+            sa.Column("id", sa.String(64), primary_key=True),
+            sa.Column("user_id", sa.String(64), nullable=False),
+            sa.Column("agent_id", sa.String(64), nullable=False),
+            sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
+            sa.UniqueConstraint("user_id", "agent_id", name="uq_agent_favorites_user_agent"),
+        )
+        op.create_index("ix_agent_favorites_user_id", "agent_favorites", ["user_id"])
+        op.create_index("ix_agent_favorites_agent_id", "agent_favorites", ["agent_id"])
+
+
+def downgrade() -> None:
+    if _has_table("agent_favorites"):
+        op.drop_table("agent_favorites")

offline-backend-20260512/backend/packages/harness/deerflow/persistence/migrations/versions/20260528_02_user_custom_prompts.py

@@ -0,0 +1,45 @@
+"""Add ``user_custom_prompts`` for per-user saved prompts (常用提示词).
+
+Bound to the user (not the browser) so prompts follow the account.
+
+Revision ID: 20260528_02
+Revises: 20260528_01
+Create Date: 2026-05-28
+"""
+
+from __future__ import annotations
+
+import sqlalchemy as sa
+from alembic import op
+from sqlalchemy import inspect
+
+
+revision = "20260528_02"
+down_revision = "20260528_01"
+branch_labels = None
+depends_on = None
+
+
+def _has_table(table_name: str) -> bool:
+    inspector = inspect(op.get_bind())
+    return table_name in inspector.get_table_names()
+
+
+def upgrade() -> None:
+    if not _has_table("user_custom_prompts"):
+        op.create_table(
+            "user_custom_prompts",
+            sa.Column("id", sa.String(64), primary_key=True),
+            sa.Column("user_id", sa.String(64), nullable=False),
+            sa.Column("title", sa.String(255), nullable=False, server_default=""),
+            sa.Column("content", sa.Text(), nullable=False),
+            sa.Column("sort_order", sa.Integer(), nullable=False, server_default="0"),
+            sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
+            sa.Column("updated_at", sa.DateTime(timezone=True), nullable=False),
+        )
+        op.create_index("ix_user_custom_prompts_user_id", "user_custom_prompts", ["user_id"])
+
+
+def downgrade() -> None:
+    if _has_table("user_custom_prompts"):
+        op.drop_table("user_custom_prompts")

offline-backend-20260512/backend/packages/harness/deerflow/persistence/models/__init__.py

@@ -15,7 +15,7 @@ its storage implementation lives in ``deerflow.runtime.events.store.db`` and
 there is no matching entity directory.
 """
 
-from deerflow.persistence.agents.model import AgentRow
+from deerflow.persistence.agents.model import AgentFavoriteRow, AgentRow
 from deerflow.persistence.memory_config.model import UserMemoryConfigRow
 from deerflow.persistence.feedback.model import FeedbackRow
 from deerflow.persistence.llm_metrics.model import LlmCallMetricRow
@@ -36,9 +36,11 @@ from deerflow.persistence.thread_meta.model import ThreadMetaRow
 from deerflow.persistence.thread_shares.model import ThreadShareRow
 from deerflow.persistence.tool_metrics.model import ToolCallMetricRow
 from deerflow.persistence.user.model import UserRow
+from deerflow.persistence.user_prompts.model import UserCustomPromptRow
 
 __all__ = [
     "FeedbackRow",
+    "AgentFavoriteRow",
     "AgentRow",
     "LlmCallMetricRow",
     "NotificationRow",
@@ -57,6 +59,7 @@ __all__ = [
     "ThreadMetaRow",
     "ThreadShareRow",
     "ToolCallMetricRow",
+    "UserCustomPromptRow",
     "UserMemoryConfigRow",
     "UserRow",
 ]

offline-backend-20260512/backend/packages/harness/deerflow/persistence/user_prompts/__init__.py

@@ -0,0 +1,18 @@
+"""Per-user custom prompt persistence."""
+
+from __future__ import annotations
+
+from deerflow.persistence.user_prompts.model import UserCustomPromptRow
+from deerflow.persistence.user_prompts.sql import UserCustomPromptRepository
+
+__all__ = [
+    "UserCustomPromptRepository",
+    "UserCustomPromptRow",
+    "make_user_prompt_store",
+]
+
+
+def make_user_prompt_store(session_factory):
+    if session_factory is None:
+        return None
+    return UserCustomPromptRepository(session_factory)

offline-backend-20260512/backend/packages/harness/deerflow/persistence/user_prompts/model.py

@@ -0,0 +1,23 @@
+"""ORM model for per-user custom prompts (常用提示词)."""
+
+from __future__ import annotations
+
+from datetime import UTC, datetime
+
+from sqlalchemy import Integer, String, Text
+from sqlalchemy.orm import Mapped, mapped_column
+
+from deerflow.persistence.base import Base
+from deerflow.persistence.types import BeijingDateTime
+
+
+class UserCustomPromptRow(Base):
+    __tablename__ = "user_custom_prompts"
+
+    id: Mapped[str] = mapped_column(String(64), primary_key=True)
+    user_id: Mapped[str] = mapped_column(String(64), nullable=False, index=True)
+    title: Mapped[str] = mapped_column(String(255), nullable=False, default="")
+    content: Mapped[str] = mapped_column(Text, nullable=False)
+    sort_order: Mapped[int] = mapped_column(Integer, nullable=False, default=0)
+    created_at: Mapped[datetime] = mapped_column(BeijingDateTime(), nullable=False, default=lambda: datetime.now(UTC))
+    updated_at: Mapped[datetime] = mapped_column(BeijingDateTime(), nullable=False, default=lambda: datetime.now(UTC), onupdate=lambda: datetime.now(UTC))

offline-backend-20260512/backend/packages/harness/deerflow/persistence/user_prompts/sql.py

@@ -0,0 +1,74 @@
+"""SQLAlchemy-backed per-user custom prompt repository."""
+
+from __future__ import annotations
+
+from datetime import UTC, datetime
+from typing import Any
+
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession, async_sessionmaker
+
+from deerflow.persistence.user_prompts.model import UserCustomPromptRow
+
+
+def _row_to_dict(row: UserCustomPromptRow) -> dict[str, Any]:
+    data = row.to_dict()
+    for key in ("created_at", "updated_at"):
+        value = data.get(key)
+        if isinstance(value, datetime):
+            data[key] = value.isoformat()
+    return data
+
+
+class UserCustomPromptRepository:
+    def __init__(self, session_factory: async_sessionmaker[AsyncSession]) -> None:
+        self._sf = session_factory
+
+    async def list_for_user(self, user_id: str) -> list[dict[str, Any]]:
+        stmt = (
+            select(UserCustomPromptRow)
+            .where(UserCustomPromptRow.user_id == user_id)
+            .order_by(UserCustomPromptRow.sort_order.asc(), UserCustomPromptRow.created_at.asc())
+        )
+        async with self._sf() as session:
+            result = await session.execute(stmt)
+            return [_row_to_dict(row) for row in result.scalars()]
+
+    async def create(self, user_id: str, data: dict[str, Any]) -> dict[str, Any]:
+        now = datetime.now(UTC)
+        row = UserCustomPromptRow(
+            id=data["id"],
+            user_id=user_id,
+            title=data.get("title") or "",
+            content=data["content"],
+            sort_order=data.get("sort_order", 0),
+            created_at=now,
+            updated_at=now,
+        )
+        async with self._sf() as session:
+            session.add(row)
+            await session.commit()
+            await session.refresh(row)
+            return _row_to_dict(row)
+
+    async def update(self, prompt_id: str, user_id: str, data: dict[str, Any]) -> dict[str, Any] | None:
+        async with self._sf() as session:
+            row = await session.get(UserCustomPromptRow, prompt_id)
+            if row is None or row.user_id != user_id:
+                return None
+            for key in ("title", "content", "sort_order"):
+                if key in data:
+                    setattr(row, key, data[key])
+            row.updated_at = datetime.now(UTC)
+            await session.commit()
+            await session.refresh(row)
+            return _row_to_dict(row)
+
+    async def delete(self, prompt_id: str, user_id: str) -> bool:
+        async with self._sf() as session:
+            row = await session.get(UserCustomPromptRow, prompt_id)
+            if row is None or row.user_id != user_id:
+                return False
+            await session.delete(row)
+            await session.commit()
+            return True

offline-backend-20260512/backend/packages/harness/deerflow/runtime/references.py

@@ -60,7 +60,7 @@ def build_latest_reference_prompt(messages: list[Any]) -> tuple[str, dict[str, A
         title = source.get("title") or f"参考文献 {source.get('index')}"
         meta = " · ".join(
             str(item)
-            for item in (source.get("source"), source.get("time"))
+            for item in (source.get("source"), source.get("author"), source.get("time"))
             if item
         )
         content = str(source.get("content") or source.get("snippet") or "").strip()
@@ -81,6 +81,7 @@ def build_latest_reference_prompt(messages: list[Any]) -> tuple[str, dict[str, A
             "3. 只能引用上方存在的编号，不要编造编号。",
             "4. 不要使用工具原始编号，也不要重新排序编号。",
             "5. 不要在正文末尾堆完整参考文献列表，前端会展示右侧参考文献。",
+            "6. 参考文献模式下不要输出 <skill-display .../> 占位符，也不要在正文里额外渲染卡片、表格或列表；前端会在右侧参考文献区域展示。",
         ]
     )
     return "\n".join(lines).strip(), batch
@@ -157,6 +158,26 @@ def build_reference_debug_summary(messages: list[Any], batches: list[dict[str, A
     }
 
 
+def build_reference_debug_payload(
+    *,
+    messages: list[Any] | None = None,
+    batch: dict[str, Any] | None = None,
+    batches: list[dict[str, Any]] | None = None,
+    prompt: str | None = None,
+) -> dict[str, Any]:
+    """Return a JSON-safe payload for comparing model prompts and UI batches."""
+    payload: dict[str, Any] = {}
+    if messages is not None:
+        payload["message_trace"] = [_message_debug_item(index, message) for index, message in enumerate(messages)]
+    if prompt is not None:
+        payload["model_reference_prompt"] = prompt
+    if batch is not None:
+        payload["batch"] = _batch_debug_item(batch)
+    if batches is not None:
+        payload["batches"] = [_batch_debug_item(item) for item in batches]
+    return payload
+
+
 def adapt_reference_batches_with_displays(
     batches: list[dict[str, Any]],
     displays: dict[str, SkillDisplayConfig | dict[str, Any]],
@@ -235,6 +256,65 @@ def adapt_reference_batches_with_displays(
     return adapted_batches
 
 
+def _message_debug_item(index: int, message: Any) -> dict[str, Any]:
+    content = _message_content(message)
+    tool_calls = _extract_tool_calls(message)
+    return {
+        "position": index,
+        "type": _message_type(message),
+        "id": _message_id(message),
+        "name": _message_name(message) if _message_type(message) == "tool" else None,
+        "tool_call_id": _message_tool_call_id(message) if _message_type(message) == "tool" else None,
+        "tool_calls": [
+            {
+                "id": call.get("id"),
+                "name": call.get("name"),
+                "args": call.get("args"),
+            }
+            for call in tool_calls
+        ],
+        "content_length": len(content) if isinstance(content, str) else None,
+    }
+
+
+def _batch_debug_item(batch: dict[str, Any]) -> dict[str, Any]:
+    sources = [source for source in batch.get("sources", []) if isinstance(source, dict)]
+    return {
+        "id": batch.get("id"),
+        "assistant_message_id": batch.get("assistant_message_id"),
+        "source_count": len(sources),
+        "sources": [_source_debug_item(source) for source in sources],
+    }
+
+
+def _source_debug_item(source: dict[str, Any]) -> dict[str, Any]:
+    raw = source.get("raw")
+    raw_record = raw if isinstance(raw, dict) else {}
+    return {
+        "index": source.get("index"),
+        "id": source.get("id"),
+        "skillName": source.get("skillName"),
+        "mode": source.get("mode"),
+        "title": source.get("title"),
+        "source": source.get("source"),
+        "author": source.get("author"),
+        "time": source.get("time"),
+        "url": source.get("url"),
+        "snippet": source.get("snippet"),
+        "content": source.get("content"),
+        "raw_id_candidates": {
+            "recUuid": raw_record.get("recUuid"),
+            "rec_uuid": raw_record.get("rec_uuid"),
+            "uuid": raw_record.get("uuid"),
+            "docId": raw_record.get("docId"),
+            "doc_id": raw_record.get("doc_id"),
+            "id": raw_record.get("id"),
+            "url": raw_record.get("url"),
+        },
+        "raw": raw,
+    }
+
+
 def _collect_sources(messages: list[Any]) -> list[dict[str, Any]]:
     sources: list[dict[str, Any]] = []
     pending_tool_contexts: list[dict[str, Any]] = []
@@ -536,6 +616,10 @@ def _pick(layers: list[dict[str, Any]], keys: tuple[str, ...]) -> str | None:
             value = layer.get(key)
             if isinstance(value, (str, int, float)) and str(value).strip():
                 return str(value).strip()
+            if isinstance(value, list):
+                text = _stringify(value, "")
+                if text:
+                    return text
     return None
 
 
@@ -604,6 +688,7 @@ def _normalize_with_display_config(
     snippet = _stringify(_get_by_path(raw, display.summary_field), "")
     content = _stringify(_get_by_path(raw, display.content_field), snippet)
     source = _stringify(_get_by_path(raw, display.source_field), "") or None
+    author = _stringify(_get_by_path(raw, display.author_field), "") or None
     item_type = _stringify(_get_by_path(raw, display.type_field), "result") or "result"
     time = _stringify(_get_by_path(raw, display.time_field), "") or None
     item_id = _stringify(_get_by_path(raw, display.id_field), "") or None
@@ -620,6 +705,7 @@ def _normalize_with_display_config(
         "snippet": snippet,
         "content": content,
         "source": source,
+        "author": author,
         "time": time,
         "url": url,
         "score": score,
@@ -667,6 +753,13 @@ def _stringify(value: Any, fallback: str = "") -> str:
     if isinstance(value, (str, int, float)):
         text = str(value).strip()
         return text if text else fallback
+    if isinstance(value, list):
+        parts = [
+            str(item).strip()
+            for item in value
+            if isinstance(item, (str, int, float)) and str(item).strip()
+        ]
+        return "、".join(parts) if parts else fallback
     return fallback
 
 
@@ -705,6 +798,7 @@ def _normalize_reference_item(
     content = _pick(layers, ("content", "page_content", "summary", "snippet", "description", "content_preview", "abstract", "body"))
     url = _pick(layers, ("url", "link", "href", "source_url", "web_url"))
     source = _pick(layers, ("source", "source1", "site", "provider", "from"))
+    author = _pick(layers, ("author", "authors", "creator", "writer", "byline", "publisher", "owner"))
     time = _pick(layers, ("time", "date", "published_date", "publishedDate", "m_publish", "publish_time", "created_at"))
     item_id = _pick(layers, ("recUuid", "rec_uuid", "uuid", "docId", "doc_id", "id", "recordId", "taskId"))
     score_raw = _pick(layers, ("score", "relevance", "similarity"))
@@ -729,6 +823,7 @@ def _normalize_reference_item(
         "snippet": content,
         "content": content,
         "source": source,
+        "author": author,
         "time": time,
         "url": url,
         "score": score,

offline-backend-20260512/backend/packages/harness/deerflow/runtime/scheduler/__init__.py

@@ -1,4 +1,10 @@
 from .registry import get_scheduled_task_service, set_scheduled_task_service
-from .service import DuplicateTaskNameError, ScheduledTaskService
+from .service import DuplicateTaskNameError, MarkdownRequirementError, ScheduledTaskService
 
-__all__ = ["DuplicateTaskNameError", "ScheduledTaskService", "get_scheduled_task_service", "set_scheduled_task_service"]
+__all__ = [
+    "DuplicateTaskNameError",
+    "MarkdownRequirementError",
+    "ScheduledTaskService",
+    "get_scheduled_task_service",
+    "set_scheduled_task_service",
+]

offline-backend-20260512/backend/packages/harness/deerflow/runtime/scheduler/service.py

@@ -36,6 +36,27 @@ _RUN_ATTEMPT_TIMEOUT_SECONDS = 30 * 60  # 30 minutes
 # Retries allowed after the first attempt before the run is marked "failed".
 _MAX_RUN_RETRIES = 3
 
+# Minimum length (in characters) the generated Markdown file's content must
+# reach when a task has ``require_markdown`` enabled. The model is NOT told about
+# this threshold — it is enforced silently by the service.
+_MARKDOWN_MIN_CHARS = 100
+
+# Instruction appended to the prompt when ``require_markdown`` is on. It only
+# asks the model to produce a ``.md`` file; the length requirement is enforced
+# in code (not surfaced to the model).
+_MARKDOWN_INSTRUCTION = (
+    "\n\n【输出要求】请将本次任务的最终结果生成为一个 Markdown(.md) 格式的文件，"
+    "保存到 /mnt/user-data/outputs 目录，并使用 present_files 展示该文件。"
+)
+
+
+class MarkdownRequirementError(Exception):
+    """Raised when a ``require_markdown`` task did not produce a valid .md file.
+
+    The run is marked failed (no retry) so the user can see the output did not
+    meet the requirement (a Markdown file longer than the minimum length).
+    """
+
 
 class DuplicateTaskNameError(Exception):
     """A user already owns a scheduled task with the requested name.
@@ -166,6 +187,42 @@ def _looks_like_markdown(content: str) -> bool:
     return any(re.search(pattern, content, flags=re.MULTILINE) for pattern in markdown_patterns)
 
 
+def _task_requires_markdown(task: dict[str, Any]) -> bool:
+    """Whether a task has the ``require_markdown`` flag set.
+
+    The flag lives inside ``execution_context`` (no dedicated column) so the
+    feature works without a schema migration on existing databases.
+    """
+    ctx = task.get("execution_context")
+    return bool(isinstance(ctx, dict) and ctx.get("require_markdown"))
+
+
+def _has_valid_markdown_file(result_files: list[dict[str, Any]] | None) -> bool:
+    """True when the run produced a ``.md`` file whose text exceeds the minimum.
+
+    For tasks with ``require_markdown`` enabled the run only succeeds if at least
+    one generated file is Markdown (``.md``/``.markdown``) and its content is
+    longer than :data:`_MARKDOWN_MIN_CHARS` characters. The file is read from its
+    archived path; binary/undecodable files never qualify.
+    """
+    for item in result_files or []:
+        if not isinstance(item, dict):
+            continue
+        name = str(item.get("name") or "")
+        if not name.lower().endswith((".md", ".markdown")):
+            continue
+        path = item.get("path")
+        if not isinstance(path, str) or not path:
+            continue
+        try:
+            text = Path(path).read_text(encoding="utf-8", errors="ignore")
+        except Exception:
+            continue
+        if len(text.strip()) > _MARKDOWN_MIN_CHARS:
+            return True
+    return False
+
+
 def _build_result(
     task: dict[str, Any],
     *,
@@ -451,6 +508,7 @@ class ScheduledTaskService:
         source_agent_name: str | None = None,
         execution_agent_name: str | None = None,
         execution_context: dict[str, Any] | None = None,
+        require_markdown: bool = False,
     ) -> dict[str, Any]:
         tz = _normalize_timezone(timezone)
         cron_expr = _schedule_to_cron(schedule)
@@ -461,6 +519,7 @@ class ScheduledTaskService:
         context = dict(execution_context or {})
         if execution_agent_name:
             context["agent_name"] = execution_agent_name
+        context["require_markdown"] = bool(require_markdown)
         normalized_description = description.strip() if isinstance(description, str) else None
         task = await self.store.create_task(
             {
@@ -519,6 +578,18 @@ class ScheduledTaskService:
             update_data["next_run_at"] = _next_run_at(cron_expr, tz)
         if data.get("execution_context") is not None:
             update_data["execution_context"] = data["execution_context"]
+        # ``require_markdown`` is persisted inside execution_context (no dedicated
+        # column), so merge it into the context being written — preferring a
+        # context supplied in the same request, else the task's existing one.
+        if data.get("require_markdown") is not None:
+            base_ctx = (
+                update_data["execution_context"]
+                if "execution_context" in update_data
+                else (existing or {}).get("execution_context")
+            )
+            ctx = dict(base_ctx or {})
+            ctx["require_markdown"] = bool(data["require_markdown"])
+            update_data["execution_context"] = ctx
         # Re-enabling a task must also schedule it forward from now, otherwise a
         # paused task with a stale/empty next_run_at would never fire again.
         if update_data.get("enabled") is True and "next_run_at" not in update_data:
@@ -831,6 +902,10 @@ class ScheduledTaskService:
                 run_id=run_id,
                 output_snapshot=output_snapshot,
             )
+        if _task_requires_markdown(task) and not _has_valid_markdown_file(result_files):
+            raise MarkdownRequirementError(
+                f"未满足输出要求：需生成 Markdown(.md) 文件，且文件正文不少于 {_MARKDOWN_MIN_CHARS} 字"
+            )
         return _build_result(task, scheduled_for=scheduled_for, content=result_text, files=result_files)
 
     async def _ensure_thread_meta(self, user_id: str, thread_id: str) -> None:
@@ -895,6 +970,8 @@ class ScheduledTaskService:
                 f"计划时间：{scheduled_for.astimezone(ZoneInfo(task.get('timezone') or 'UTC')).isoformat()}\n\n"
                 f"{task['prompt']}"
             )
+            if execution_context.get("require_markdown"):
+                content += _MARKDOWN_INSTRUCTION
             body = RunCreateRequest(
                 assistant_id="lead_agent",
                 input={"messages": [{"type": "human", "content": content}]},