Private/issue 1096 api latency logging

Makefile

@@ -214,9 +214,12 @@ evaluate:
 ##################################################
 
 # OpenAPI and model generation
-.PHONY: merge-openapi generate-models generate-frontend-sdk
+.PHONY: openapi-check merge-openapi generate-models generate-frontend-sdk
+openapi-check:
+	@npx --yes @redocly/cli bundle aperag/api/openapi.yaml --output /tmp/openapi-check-bundle.yaml && rm -f /tmp/openapi-check-bundle.yaml
+
 merge-openapi:
-	@cd aperag && redocly bundle ./api/openapi.yaml > ./api/openapi.merged.yaml
+	@cd aperag && npx --yes @redocly/cli bundle ./api/openapi.yaml > ./api/openapi.merged.yaml
 
 generate-models: merge-openapi
 	@datamodel-codegen \

PR_DESCRIPTION_1096.md

@@ -0,0 +1,17 @@
+# Summary
+This PR updates marketplace collection graph pages to use marketplace-safe graph APIs and improves graph page layout behavior on shorter screens.
+
+## What changed
+- Render the marketplace collection `/graph` page with the new graph-hybrid view.
+- Add read-only marketplace graph endpoints for `embedding-map` and `entity-search` so published marketplace collections do not call workspace-only graph APIs.
+- Let `CollectionGraphHybrid` switch between workspace and marketplace data sources.
+- Give graph pages a 720px minimum graph area so shorter screens can scroll instead of compressing the canvas.
+
+## Why
+Marketplace collection graph pages should not depend on workspace-only APIs. This change introduces marketplace-safe read-only graph endpoints and updates frontend routing/data-source behavior so graph rendering works correctly for published collections while preserving workspace behavior.
+
+## Validation
+- `make openapi-check`
+- `uv run ruff check aperag/domains/marketplace/api/routes.py`
+- `yarn type-check`
+- `yarn lint` (passes with pre-existing unrelated warnings)

PR_DESCRIPTION_1963.md

@@ -0,0 +1,46 @@
+# Summary
+This PR addresses issue #1963 where `PUT /api/v2/collections/{id}` returns stable `500 DATABASE_ERROR` when updating an existing collection config in apemind POC (SG) / SG evaluation environments.
+
+## Problem
+- Updating an existing `type=document` collection via `PUT /api/v2/collections/{collection_id}` consistently fails with:
+  - `{"success":false,"error_code":"DATABASE_ERROR","code":1050,"message":"数据库出现错误，请稍后重试。"}`
+- Retry does not recover; failure is stable.
+
+## Reproduction
+1. Create a `type=document` collection (POST path succeeds).
+2. Update config with `PUT /api/v2/collections/{collection_id}`:
+   - Reproduces when changing only `enable_vector` / `fulltext` / `embedding`.
+   - Also reproduces when enabling knowledge graph (`enable_knowledge_graph=true`).
+3. Observe stable `500 DATABASE_ERROR`.
+
+## Impact
+- Existing collections cannot be reconfigured.
+- Typical operation "enable knowledge graph after collection creation" is blocked.
+- Current workaround is delete + recreate collection, which is high cost and may lose built index state.
+
+## Isolation Findings
+- `POST /api/v2/collections` (CREATE): normal.
+- `GET` paths: normal.
+- Only `PUT` update path is failing.
+- Failure is not knowledge-graph specific.
+
+## Scope in this PR
+- Triage and root-cause analysis for collection update failure in update path.
+- Confirm ownership boundary between KB domain collection update service and DB layer.
+- Implement and validate fix for `DATABASE_ERROR` in update flow.
+
+## Validation Plan
+- Reproduce on apemind POC (SG) with an existing collection.
+- Verify PUT update succeeds for:
+  - Non-graph config-only changes (`enable_vector` / `fulltext` / `embedding`).
+  - Graph enablement path (`enable_knowledge_graph=true`).
+- Regression check:
+  - CREATE remains normal.
+  - GET remains normal.
+  - No regression in collection update behavior across existing test fixtures.
+
+## Context
+- Environment where issue was found: apemind POC (SG).
+- Discovery date: 2026-07-01.
+- Reporter: @cuiwenbo (崔文博), during `task feat: chat #17`.
+- Tracking issue: #1963.

README-zh.md

@@ -6,8 +6,6 @@
 
 ApeRAG 是一个生产级 RAG（检索增强生成）平台，结合了图 RAG、向量搜索、全文搜索和先进的 AI 智能体。构建具有混合检索、多模态文档处理、智能代理和企业级管理功能的复杂 AI 应用程序。
 
-**🚀 [在线体验 ApeRAG](https://rag.apecloud.com/)** - 通过我们的托管演示体验完整的平台功能
-
 ApeRAG 是你构建自己的知识图谱、进行上下文工程以及部署能够自主搜索和推理知识库的智能 AI 代理的最佳选择。
 
 [Read English Documentation](README.md)
@@ -51,7 +49,7 @@ ApeRAG 支持 [MCP（模型上下文协议）](https://modelcontextprotocol.io/)
 {
   "mcpServers": {
     "aperag-mcp": {
-      "url": "https://rag.apecloud.com/mcp/",
+      "url": "http://localhost:8000/mcp/",
       "headers": {
         "Authorization": "Bearer your-api-key-here"
       }
@@ -64,7 +62,7 @@ ApeRAG 支持 [MCP（模型上下文协议）](https://modelcontextprotocol.io/)
 1. **HTTP Authorization 头**（推荐）：`Authorization: Bearer your-api-key`
 2. **环境变量**（备用）：`APERAG_API_KEY=your-api-key`
 
-**重要提示**：将 `https://rag.apecloud.com` 替换为您实际的 ApeRAG API 地址，将 `your-api-key-here` 替换为 ApeRAG 设置中的有效 API 密钥。
+**重要提示**：若为非本机部署，请将示例 URL 换成实际 API 源地址对应的 MCP 路径（如 `https://<你的域名>/mcp/`）。将 `your-api-key-here` 替换为 ApeRAG 设置中的有效 API 密钥。
 
 MCP 服务器提供：
 - **集合浏览**：列出和探索您的知识集合

README.md

@@ -1,9 +1,6 @@
 # ApeRAG
 [![Trust Score](https://archestra.ai/mcp-catalog/api/badge/quality/apecloud/ApeRAG)](https://archestra.ai/mcp-catalog/apecloud__aperag)
 
-**🚀 [Try ApeRAG Live Demo](https://rag.apecloud.com/)** - Experience the full platform capabilities with our hosted demo
-
-
 ![HarryPotterKG2.png](docs%2Fen-US%2Fimages%2FHarryPotterKG2.png)
 
 ![chat2.png](docs%2Fen-US%2Fimages%2Fchat2.png)
@@ -52,7 +49,7 @@ ApeRAG supports [MCP (Model Context Protocol)](https://modelcontextprotocol.io/)
 {
   "mcpServers": {
     "aperag-mcp": {
-      "url": "https://rag.apecloud.com/mcp/",
+      "url": "http://localhost:8000/mcp/",
       "headers": {
         "Authorization": "Bearer your-api-key-here"
       }
@@ -65,7 +62,7 @@ ApeRAG supports [MCP (Model Context Protocol)](https://modelcontextprotocol.io/)
 1. **HTTP Authorization Header** (Recommended): `Authorization: Bearer your-api-key`
 2. **Environment Variable** (Fallback): `APERAG_API_KEY=your-api-key`
 
-**Important**: Replace `https://rag.apecloud.com` with your actual ApeRAG API URL and `your-api-key-here` with a valid API key from your ApeRAG settings.
+**Important**: Use your deployed API origin if not local (e.g. `https://your-host/mcp/`). Replace `your-api-key-here` with a valid API key from your ApeRAG settings.
 
 The MCP server provides:
 - **Collection browsing**: List and explore your knowledge collections

aperag/api/components/schemas/document.yaml

@@ -251,3 +251,89 @@ confirmDocumentsResponse:
   required:
     - confirmed_count
     - failed_count
+
+fetchUrlRequest:
+  type: object
+  properties:
+    urls:
+      type: array
+      items:
+        type: string
+        format: uri
+      minItems: 1
+      maxItems: 10
+      description: List of URLs to fetch and import (max 10)
+      example:
+        - "https://example.com/article1"
+        - "https://example.com/article2"
+  required:
+    - urls
+
+fetchUrlResultItem:
+  type: object
+  properties:
+    url:
+      type: string
+      description: The source URL
+    fetch_status:
+      type: string
+      enum:
+        - success
+        - error
+      description: Whether the URL was fetched successfully
+    document_id:
+      type: string
+      description: ID of the created document (only present on success)
+    filename:
+      type: string
+      description: Filename of the created document (only present on success)
+    size:
+      type: integer
+      description: Size of the created document in bytes (only present on success)
+    status:
+      type: string
+      description: Document status (only present on success)
+    error:
+      type: string
+      description: Error message (only present on failure)
+  required:
+    - url
+    - fetch_status
+
+fetchUrlResponse:
+  type: object
+  properties:
+    results:
+      type: array
+      items:
+        $ref: '#/fetchUrlResultItem'
+      description: Results for each URL
+    total:
+      type: integer
+      description: Total number of URLs processed
+    succeeded:
+      type: integer
+      description: Number of URLs successfully fetched
+    failed:
+      type: integer
+      description: Number of URLs that failed
+  required:
+    - results
+    - total
+    - succeeded
+    - failed
+
+stagedDocumentsResponse:
+  type: object
+  properties:
+    documents:
+      type: array
+      items:
+        $ref: '#/uploadDocumentResponse'
+      description: List of staged (UPLOADED) documents awaiting confirmation
+    total:
+      type: integer
+      description: Total number of staged documents
+  required:
+    - documents
+    - total

aperag/api/openapi.yaml

@@ -73,6 +73,10 @@ paths:
     $ref: './paths/collections.yaml#/upload_document'
   /collections/{collection_id}/documents/confirm:
     $ref: './paths/collections.yaml#/confirm_documents'
+  /collections/{collection_id}/documents/fetch-url:
+    $ref: './paths/collections.yaml#/fetch_url_document'
+  /collections/{collection_id}/documents/staged:
+    $ref: './paths/collections.yaml#/list_staged_documents'
   /collections/{collection_id}/searches:
     $ref: './paths/collections.yaml#/searches'
   /collections/{collection_id}/searches/{search_id}:
@@ -111,6 +115,10 @@ paths:
     $ref: './paths/collections.yaml#/graph_suggestion_action'
   /collections/{collection_id}/graphs/export/kg-eval:
     $ref: './paths/collections.yaml#/graph_export_kg_eval'
+  /collections/{collection_id}/graphs/embedding-map:
+    $ref: './paths/collections.yaml#/graph_embedding_map'
+  /collections/{collection_id}/graphs/entity-search:
+    $ref: './paths/collections.yaml#/graph_entity_search'
   /collections/{collection_id}/sharing:
     $ref: './paths/collections.yaml#/sharing'
 
@@ -131,6 +139,10 @@ paths:
     $ref: './paths/marketplace.yaml#/marketplaceCollectionDocumentObject'
   /marketplace/collections/{collection_id}/graph:
     $ref: './paths/marketplace.yaml#/marketplaceCollectionGraph'
+  /marketplace/collections/{collection_id}/graph/embedding-map:
+    $ref: './paths/marketplace.yaml#/marketplaceCollectionGraphEmbeddingMap'
+  /marketplace/collections/{collection_id}/graph/entity-search:
+    $ref: './paths/marketplace.yaml#/marketplaceCollectionGraphEntitySearch'
 
   # apikeys
   /apikeys: