✨ file to text tool

Author: Zhi-a

.gitignore

@@ -1,5 +1,4 @@
 .idea
-.gitignore
 /.env
 .vscode
 

backend/agents/create_agent_info.py

@@ -18,6 +18,7 @@
 from services.tenant_config_service import get_selected_knowledge_list
 from services.remote_mcp_service import get_remote_mcp_server_list
 from services.memory_config_service import build_memory_context
+from services.image_service import get_vlm_model
 from database.agent_db import search_agent_info_by_agent_id, query_sub_agents_id_list
 from database.tool_db import search_tools_for_sub_agent
 from database.model_management_db import get_model_records, get_model_by_model_id
@@ -244,6 +245,12 @@ async def create_tool_config_list(agent_id, tenant_id, user_id):
                 "storage_client": minio_client,
                 "data_process_service_url": DATA_PROCESS_SERVICE
             }
+        elif tool_config.class_name == "AnalyzeImageTool":
+            tool_config.metadata = {
+                "vlm_model": get_vlm_model(tenant_id=tenant_id),
+                "storage_client": minio_client,
+            }
+
         tool_config_list.append(tool_config)
 
     return tool_config_list

backend/apps/agent_app.py

@@ -134,7 +134,11 @@ async def import_agent_api(request: AgentImportRequest, authorization: Optional[
     import an agent
     """
     try:
-        await import_agent_impl(request.agent_info, authorization)
+        await import_agent_impl(
+            request.agent_info,
+            authorization,
+            force_import=request.force_import
+        )
         return {}
     except Exception as e:
         logger.error(f"Agent import error: {str(e)}")

backend/apps/vectordatabase_app.py

@@ -5,7 +5,7 @@
 from fastapi import APIRouter, Body, Depends, Header, HTTPException, Path, Query
 from fastapi.responses import JSONResponse
 
-from consts.model import IndexingResponse
+from consts.model import ChunkCreateRequest, ChunkUpdateRequest, HybridSearchRequest, IndexingResponse
 from nexent.vector_database.base import VectorDatabaseCore
 from services.vectordatabase_service import (
     ElasticSearchService,
@@ -226,3 +226,125 @@ def get_index_chunks(
             f"Error getting chunks for index '{index_name}': {error_msg}")
         raise HTTPException(
             status_code=HTTPStatus.INTERNAL_SERVER_ERROR, detail=f"Error getting chunks: {error_msg}")
+
+
+@router.post("/{index_name}/chunk")
+def create_chunk(
+        index_name: str = Path(..., description="Name of the index"),
+        payload: ChunkCreateRequest = Body(..., description="Chunk data"),
+        vdb_core: VectorDatabaseCore = Depends(get_vector_db_core),
+        authorization: Optional[str] = Header(None),
+):
+    """Create a manual chunk."""
+    try:
+        user_id, _ = get_current_user_id(authorization)
+        result = ElasticSearchService.create_chunk(
+            index_name=index_name,
+            chunk_request=payload,
+            vdb_core=vdb_core,
+            user_id=user_id,
+        )
+        return JSONResponse(status_code=HTTPStatus.OK, content=result)
+    except Exception as exc:
+        logger.error(
+            "Error creating chunk for index %s: %s", index_name, exc, exc_info=True
+        )
+        raise HTTPException(
+            status_code=HTTPStatus.INTERNAL_SERVER_ERROR, detail=str(exc)
+        )
+
+
+@router.put("/{index_name}/chunk/{chunk_id}")
+def update_chunk(
+        index_name: str = Path(..., description="Name of the index"),
+        chunk_id: str = Path(..., description="Chunk identifier"),
+        payload: ChunkUpdateRequest = Body(...,
+                                           description="Chunk update payload"),
+        vdb_core: VectorDatabaseCore = Depends(get_vector_db_core),
+        authorization: Optional[str] = Header(None),
+):
+    """Update an existing chunk."""
+    try:
+        user_id, _ = get_current_user_id(authorization)
+        result = ElasticSearchService.update_chunk(
+            index_name=index_name,
+            chunk_id=chunk_id,
+            chunk_request=payload,
+            vdb_core=vdb_core,
+            user_id=user_id,
+        )
+        return JSONResponse(status_code=HTTPStatus.OK, content=result)
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=HTTPStatus.BAD_REQUEST, detail=str(exc))
+    except Exception as exc:
+        logger.error(
+            "Error updating chunk %s for index %s: %s",
+            chunk_id,
+            index_name,
+            exc,
+            exc_info=True,
+        )
+        raise HTTPException(
+            status_code=HTTPStatus.INTERNAL_SERVER_ERROR, detail=str(exc)
+        )
+
+
+@router.delete("/{index_name}/chunk/{chunk_id}")
+def delete_chunk(
+        index_name: str = Path(..., description="Name of the index"),
+        chunk_id: str = Path(..., description="Chunk identifier"),
+        vdb_core: VectorDatabaseCore = Depends(get_vector_db_core),
+        authorization: Optional[str] = Header(None),
+):
+    """Delete a chunk."""
+    try:
+        get_current_user_id(authorization)
+        result = ElasticSearchService.delete_chunk(
+            index_name=index_name,
+            chunk_id=chunk_id,
+            vdb_core=vdb_core,
+        )
+        return JSONResponse(status_code=HTTPStatus.OK, content=result)
+    except ValueError as exc:
+        raise HTTPException(status_code=HTTPStatus.NOT_FOUND, detail=str(exc))
+    except Exception as exc:
+        logger.error(
+            "Error deleting chunk %s for index %s: %s",
+            chunk_id,
+            index_name,
+            exc,
+            exc_info=True,
+        )
+        raise HTTPException(
+            status_code=HTTPStatus.INTERNAL_SERVER_ERROR, detail=str(exc)
+        )
+
+
+@router.post("/search/hybrid")
+async def hybrid_search(
+        payload: HybridSearchRequest,
+        vdb_core: VectorDatabaseCore = Depends(get_vector_db_core),
+        authorization: Optional[str] = Header(None),
+):
+    """Run a hybrid (accurate + semantic) search across indices."""
+    try:
+        _, tenant_id = get_current_user_id(authorization)
+        result = ElasticSearchService.search_hybrid(
+            index_names=payload.index_names,
+            query=payload.query,
+            tenant_id=tenant_id,
+            top_k=payload.top_k,
+            weight_accurate=payload.weight_accurate,
+            vdb_core=vdb_core,
+        )
+        return JSONResponse(status_code=HTTPStatus.OK, content=result)
+    except ValueError as exc:
+        raise HTTPException(status_code=HTTPStatus.BAD_REQUEST,
+                            detail=str(exc))
+    except Exception as exc:
+        logger.error(f"Hybrid search failed: {exc}", exc_info=True)
+        raise HTTPException(
+            status_code=HTTPStatus.INTERNAL_SERVER_ERROR,
+            detail=f"Error executing hybrid search: {str(exc)}",
+        )

backend/consts/model.py

@@ -175,6 +175,43 @@ class IndexingResponse(BaseModel):
     total_submitted: int
 
 
+class ChunkCreateRequest(BaseModel):
+    """Request payload for manual chunk creation."""
+
+    content: str = Field(..., min_length=1, description="Chunk content")
+    title: Optional[str] = Field(None, description="Optional chunk title")
+    filename: Optional[str] = Field(None, description="Associated file name")
+    path_or_url: Optional[str] = Field(None, description="Source path or URL")
+    chunk_id: Optional[str] = Field(
+        None, description="Explicit chunk identifier")
+    metadata: Dict[str, Any] = Field(
+        default_factory=dict, description="Additional chunk metadata")
+
+
+class ChunkUpdateRequest(BaseModel):
+    """Request payload for chunk updates."""
+
+    content: Optional[str] = Field(None, description="Updated chunk content")
+    title: Optional[str] = Field(None, description="Updated chunk title")
+    filename: Optional[str] = Field(None, description="Updated file name")
+    path_or_url: Optional[str] = Field(
+        None, description="Updated source path or URL")
+    metadata: Dict[str, Any] = Field(
+        default_factory=dict, description="Additional metadata updates")
+
+
+class HybridSearchRequest(BaseModel):
+    """Request payload for hybrid knowledge-base searches."""
+    query: str = Field(..., min_length=1,
+                       description="Search query text")
+    index_names: List[str] = Field(..., min_items=1,
+                                   description="List of index names to search")
+    top_k: int = Field(10, ge=1, le=100,
+                       description="Number of results to return")
+    weight_accurate: float = Field(0.5, ge=0.0, le=1.0,
+                                   description="Weight applied to accurate search scores")
+
+
 # Request models
 class ProcessParams(BaseModel):
     chunking_strategy: Optional[str] = "basic"
@@ -304,6 +341,7 @@ class ExportAndImportDataFormat(BaseModel):
 
 class AgentImportRequest(BaseModel):
     agent_info: ExportAndImportDataFormat
+    force_import: bool = False
 
 
 class ConvertStateRequest(BaseModel):

backend/prompts/utils/prompt_generate.yaml

@@ -249,3 +249,49 @@ USER_PROMPT: |-
   {% else %}
   你没有可用的助手
   {% endif %}
+
+ 
+AGENT_NAME_REGENERATE_SYSTEM_PROMPT: |-
+  ### 你是【Agent变量名调整专家】
+  你的工作是根据任务描述以及已有变量名，生成一个语义一致但不重复的 Python 变量名。
+
+  #### 约束
+  1. 变量名只能包含字母、数字和下划线，并且以字母或下划线开头，以“_assistant”结尾，长度不超过 30 个字符。
+  2. 避免与已有变量名重复，同时保持与原始名称相近的语义。
+  3. 仅输出变量名本身，不要附加额外解释或标点。
+
+
+AGENT_NAME_REGENERATE_USER_PROMPT: |-
+  ### 任务描述
+  {{ task_description }}
+
+  ### 原始变量名
+  {{ original_value }}
+
+  ### 已有变量名
+  {{ existing_values }}
+
+  请在满足约束的前提下，生成一个新的变量名，使其语义接近原始变量名但不与已有变量名重复。只输出变量名本身。
+
+
+AGENT_DISPLAY_NAME_REGENERATE_SYSTEM_PROMPT: |-
+  ### 你是【Agent展示名称调整专家】
+  你的任务是结合业务背景和已有展示名称，生成一个语义一致但不重复的 Agent 展示名称。
+
+  #### 约束
+  1. 展示名称需自然、易读，能够概括 Agent 的核心能力，以“助手”结尾，长度不超过 30 个字符。
+  2. 不得包含具体工具名称或多余符号。
+  3. 仅输出展示名称本身。
+
+
+AGENT_DISPLAY_NAME_REGENERATE_USER_PROMPT: |-
+  ### 任务描述
+  {{ task_description }}
+
+  ### 原始展示名称
+  {{ original_value }}
+
+  ### 已有展示名称
+  {{ existing_values }}
+
+  请输出一个新的展示名称，语义接近原始名称但不与已存在的展示名称重复，并满足上述约束。只输出展示名称本身。

backend/prompts/utils/prompt_generate_en.yaml

@@ -251,4 +251,50 @@ USER_PROMPT: |-
   {{assistant_description}}
   {% else %}
   You have no available assistants
-  {% endif %}
+  {% endif %}
+
+
+AGENT_NAME_REGENERATE_SYSTEM_PROMPT: |-
+  ### You are an [Agent Variable Name Refinement Expert]
+  Your job is to generate a Python-friendly variable name that keeps the same intent while avoiding duplicates.
+
+  #### Constraints
+  1. The name may contain only letters, numbers, and underscores, must start with a letter or underscore, ending with "_assistant", and stay within 30 characters.
+  2. Keep the new name semantically close to the original one while ensuring it does not duplicate existing names.
+  3. Return the variable name only—no explanations or extra symbols.
+
+
+AGENT_NAME_REGENERATE_USER_PROMPT: |-
+  ### Task Description
+  {{ task_description }}
+
+  ### Original Variable Name
+  {{ original_value }}
+
+  ### Existing Variable Names
+  {{ existing_values }}
+
+  Generate a new variable name that satisfies the constraints, keeps the original meaning, and does not duplicate any existing names. Output only the variable name.
+
+
+AGENT_DISPLAY_NAME_REGENERATE_SYSTEM_PROMPT: |-
+  ### You are an [Agent Display Name Refinement Expert]
+  You summarize the agent's capability and generate a readable display name that remains unique within the workspace.
+
+  #### Constraints
+  1. The name must be concise, easy to read, summarize the agent's responsibility, ending with "Assistant", and stay within 30 characters.
+  2. Avoid mentioning specific tool names or adding extra punctuation.
+  3. Return only the display name.
+
+
+AGENT_DISPLAY_NAME_REGENERATE_USER_PROMPT: |-
+  ### Task Description
+  {{ task_description }}
+
+  ### Original Display Name
+  {{ original_value }}
+
+  ### Existing Display Names
+  {{ existing_values }}
+
+  Output a new display name that is semantically aligned with the original while remaining unique and honoring the constraints. Return only the display name.