Pr@v3@refactor improve code

.github/workflows/build-and-push-python-pg.yml

@@ -4,9 +4,9 @@ on:
   workflow_dispatch:
     inputs:
       architecture:
-        description: 'Architecture'
+        description: "Architecture"
         required: true
-        default: 'linux/amd64'
+        default: "linux/amd64"
         type: choice
         options:
           - linux/amd64
@@ -25,15 +25,15 @@ jobs:
         run: |
           DOCKER_IMAGE=ghcr.io/1panel-dev/maxkb-base
           DOCKER_PLATFORMS=${{ github.event.inputs.architecture }}
-          TAG_NAME=python3.11-pg17.10-20260525
+          TAG_NAME=python3.13-pg17.11-20260902
           DOCKER_IMAGE_TAGS="--tag ${DOCKER_IMAGE}:${TAG_NAME}"
-          echo ::set-output name=docker_image::${DOCKER_IMAGE}
-          echo ::set-output name=version::${TAG_NAME}
-          echo ::set-output name=buildx_args::--platform ${DOCKER_PLATFORMS} --no-cache \
+          echo "docker_image=${DOCKER_IMAGE}" >> $GITHUB_OUTPUT
+          echo "version=${TAG_NAME}" >> $GITHUB_OUTPUT
+          echo "buildx_args=--platform ${DOCKER_PLATFORMS} --no-cache \
             --build-arg VERSION=${TAG_NAME} \
             --build-arg BUILD_DATE=$(date -u +'%Y-%m-%dT%H:%M:%SZ') \
             --build-arg VCS_REF=${GITHUB_SHA::8} \
-            ${DOCKER_IMAGE_TAGS} .
+            ${DOCKER_IMAGE_TAGS} ." >> $GITHUB_OUTPUT
       - name: Set up QEMU
         uses: docker/setup-qemu-action@v4
         with:
@@ -48,4 +48,4 @@ jobs:
           password: ${{ secrets.GH_TOKEN }}
       - name: Docker Buildx (build-and-push)
         run: |
-          docker buildx build --output "type=image,push=true" ${{ steps.prepare.outputs.buildx_args }} -f installer/Dockerfile-base
+          docker buildx build --output "type=image,push=true" ${{ steps.prepare.outputs.buildx_args }} -f installer/Dockerfile-base

.github/workflows/build-and-push-vector-model.yml

@@ -4,13 +4,13 @@ on:
   workflow_dispatch:
     inputs:
       dockerImageTag:
-        description: 'Docker Image Tag'
-        default: 'v2.0.3'
+        description: "Docker Image Tag"
+        default: "v2.0.3"
         required: true
       architecture:
-        description: 'Architecture'
+        description: "Architecture"
         required: true
-        default: 'linux/amd64'
+        default: "linux/amd64"
         type: choice
         options:
           - linux/amd64
@@ -32,13 +32,13 @@ jobs:
           DOCKER_PLATFORMS=${{ github.event.inputs.architecture }}
           TAG_NAME=${{ github.event.inputs.dockerImageTag }}
           DOCKER_IMAGE_TAGS="--tag ${DOCKER_IMAGE}:${TAG_NAME} --tag ${DOCKER_IMAGE}:latest"
-          echo ::set-output name=docker_image::${DOCKER_IMAGE}
-          echo ::set-output name=version::${TAG_NAME}
-          echo ::set-output name=buildx_args::--platform ${DOCKER_PLATFORMS} --no-cache \
+          echo "docker_image=${DOCKER_IMAGE}" >> $GITHUB_OUTPUT
+          echo "version=${TAG_NAME}" >> $GITHUB_OUTPUT
+          echo "buildx_args=--platform ${DOCKER_PLATFORMS} --no-cache \
             --build-arg VERSION=${TAG_NAME} \
             --build-arg BUILD_DATE=$(date -u +'%Y-%m-%dT%H:%M:%SZ') \
             --build-arg VCS_REF=${GITHUB_SHA::8} \
-            ${DOCKER_IMAGE_TAGS} .
+            ${DOCKER_IMAGE_TAGS} ." >> $GITHUB_OUTPUT
       - name: Set up QEMU
         uses: docker/setup-qemu-action@v4
         with:
@@ -53,4 +53,4 @@ jobs:
           password: ${{ secrets.GH_TOKEN }}
       - name: Docker Buildx (build-and-push)
         run: |
-          docker buildx build --output "type=image,push=true" ${{ steps.prepare.outputs.buildx_args }} -f installer/Dockerfile-vector-model
+          docker buildx build --output "type=image,push=true" ${{ steps.prepare.outputs.buildx_args }} -f installer/Dockerfile-vector-model

.github/workflows/build-and-push.yml

@@ -7,7 +7,7 @@ on:
     inputs:
       dockerImageTag:
         description: 'Image Tag'
-        default: 'v2.10.0-dev'
+        default: 'v3.0.0-dev'
         required: true
       dockerImageTagWithLatest:
         description: '是否发布latest tag（正式发版时选择，测试版本切勿选择）'

.github/workflows/python-format-check.yml

@@ -0,0 +1,47 @@
+name: Python Format Check
+
+on:
+  pull_request:
+    types: [opened, reopened]
+
+jobs:
+  pre-commit:
+    name: Ruff format changed files
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v6
+        with:
+          fetch-depth: 0
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.13"
+
+      - name: Install pre-commit
+        run: python -m pip install pre-commit==4.6.1
+
+      - name: Check changed files
+        env:
+          BASE_SHA: ${{ github.event_name == 'pull_request' && github.event.pull_request.base.sha || github.event.before }}
+          HEAD_SHA: ${{ github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}
+          DEFAULT_BRANCH: ${{ github.event.repository.default_branch }}
+        run: |
+          if ! git cat-file -e "${BASE_SHA}^{commit}" 2>/dev/null; then
+            if git show-ref --verify --quiet "refs/remotes/origin/${DEFAULT_BRANCH}"; then
+              BASE_SHA=$(git merge-base "$HEAD_SHA" "origin/${DEFAULT_BRANCH}")
+            elif git cat-file -e "${HEAD_SHA}^" 2>/dev/null; then
+              BASE_SHA=$(git rev-parse "${HEAD_SHA}^")
+            else
+              pre-commit run --all-files
+              git diff --cached --exit-code
+              exit 0
+            fi
+          fi
+
+          pre-commit run --from-ref "$BASE_SHA" --to-ref "$HEAD_SHA"
+
+          # The local hook formats and stages files. A clean index proves that
+          # every changed Python file was already formatted before CI ran.
+          git diff --cached --exit-code

.gitignore

@@ -4,6 +4,8 @@
 
 # VS Code
 .vscode
+.claude/
+.codex/
 *.project
 *.factorypath
 
@@ -190,4 +192,5 @@ apps/models_provider/impl/tencent_model_provider/model/stt.py
 tmp/
 config.yml
 .SANDBOX_BANNED_HOSTS
-copilot-instructions.md
+copilot-instructions.md
+.ai/

.pre-commit-config.yaml

@@ -0,0 +1,8 @@
+repos:
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.15.21
+    hooks:
+      - id: ruff-format
+        name: ruff format and stage
+        entry: bash -c 'ruff format "$@" && git add -- "$@"' --
+        require_serial: true

AGENTS.md

@@ -0,0 +1,123 @@
+# MaxKB-v3
+
+MaxKB (Max Knowledge Brain) is an open-source platform for building enterprise AI agents with RAG pipelines, agentic workflows, and MCP tool-use capabilities.
+
+## Agent Instructions
+
+- This file is the shared source of repository instructions for all coding agents.
+- For changes under `ui/`, read and follow `ui/AGENTS.md` before editing or reviewing frontend code.
+- Keep tool-specific entry files such as `CLAUDE.md` small; shared project rules belong here or in the applicable nested `AGENTS.md`.
+
+## Tech Stack
+
+**Backend:** Python 3.13, Django 6.0, Django REST Framework, LangChain 1.3, LangGraph 1.2, Celery 5.6, uv (package manager)
+
+**Frontend:** Vue 3.5, TypeScript, Vite 8, Element Plus, Pinia, Vue Router
+
+**Infrastructure:** PostgreSQL 17.10 + pgvector, Redis 7
+
+## Project Structure
+
+```
+apps/                    # Django backend
+  application/           # AI agents & applications
+  knowledge/             # Knowledge base & document management
+  chat/                  # Chat functionality & MCP integration
+  models_provider/       # LLM provider integrations (OpenAI, Anthropic, etc.)
+  common/                # Shared utilities, auth, caching, chunking
+  maxkb/                 # Django settings, URL routing, config
+  ops/                   # Celery task queue configuration
+  users/                 # User management
+  tools/                 # Tool/function management
+ui/                      # Vue.js frontend
+  src/
+    api/                 # API client services
+    views/               # Page views
+    components/          # Vue components
+    workflow/            # Workflow UI
+installer/               # Docker build scripts and startup scripts
+main.py                  # Application entry point
+```
+
+## Development Setup
+
+### Backend
+
+```bash
+# Install dependencies (requires Python 3.13)
+python -m uv pip install -r pyproject.toml
+
+# Run database migrations
+cd apps && python manage.py migrate
+
+# Start dev server
+python main.py dev
+# or: cd apps && python manage.py runserver 0.0.0.0:8080
+```
+
+### Frontend
+
+```bash
+cd ui
+npm install
+npm run dev        # dev server with hot reload
+npm run build      # production build
+npm run type-check # TypeScript check
+npm run lint       # ESLint
+npm run format     # Prettier
+```
+
+### Service Management
+
+```bash
+python main.py start all -d       # start all services (web + celery)
+python main.py start web -d       # web only
+python main.py start task -d      # celery worker
+python main.py start local_model -d  # local model service
+```
+
+### Docker
+
+```bash
+docker build -f installer/Dockerfile -t maxkb:latest .
+```
+
+## Configuration
+
+Key environment variables (see `.env`):
+
+| Variable                                                | Description            |
+| ------------------------------------------------------- | ---------------------- |
+| `MAXKB_DB_HOST` / `MAXKB_DB_PORT`                       | PostgreSQL connection  |
+| `MAXKB_DB_NAME` / `MAXKB_DB_USER` / `MAXKB_DB_PASSWORD` | PostgreSQL credentials |
+| `MAXKB_REDIS_HOST` / `MAXKB_REDIS_PORT`                 | Redis connection       |
+
+Settings files:
+
+- `apps/maxkb/conf.py` — main config manager (reads env vars)
+- `apps/maxkb/settings/base/web.py` — Django settings
+- `apps/maxkb/settings/lib.py` — Celery/Redis settings
+
+## Testing
+
+```bash
+cd apps && python manage.py test
+```
+
+## Code Style
+
+- Python: Ruff linter, line length 120 (`pyproject.toml`)
+- TypeScript/Vue: ESLint + Prettier (`ui/`)
+- Migrations in `apps/*/migrations/`
+
+## Key Concepts
+
+- **Knowledge Base**: Documents are chunked, embedded, and stored in PostgreSQL with pgvector for semantic search.
+- **Applications/Agents**: Built on top of knowledge bases; support workflow-based and RAG-based configurations.
+- **Models Provider**: Abstraction layer in `apps/models_provider/` supporting 10+ LLM providers via LangChain.
+- **MCP**: Model Context Protocol tools integrated via `langchain-mcp-adapters` in `apps/chat/mcp/`.
+- **Task Queue**: Heavy document processing runs async via Celery workers.
+
+## Internationalization
+
+Translations in `apps/locales/` (zh_CN, en_US, zh_Hant) and `ui/src/locales/`.

CLAUDE.md

@@ -0,0 +1,3 @@
+@AGENTS.md
+
+<!-- Shared repository instructions live in AGENTS.md. -->

CONTRIBUTING.md

@@ -14,6 +14,18 @@ This [development guideline](https://github.com/1Panel-dev/MaxKB/wiki/3-%E5%BC%8
 
 Note: If you split your pull request to small changes, please make sure any of the changes goes to master will not break anything. Otherwise, it can not be merged until this feature complete.
 
+## Code formatting
+
+This project uses [pre-commit](https://pre-commit.com/) and Ruff to format Python code. After cloning the repository and setting up the development environment, install the Git hook once:
+
+```bash
+uv sync --group dev
+uv run pre-commit install --install-hooks
+```
+
+
+The hook runs automatically on each `git commit` and stages files changed by the formatter. The same check also runs in CI for every pull request, so all formatting changes must be committed before the pull request can be merged.
+
 ## Report issues
 It is a great way to contribute by reporting an issue. Well-written and complete bug reports are always welcome! Please open an issue and follow the template to fill in required information.
 
@@ -27,4 +39,4 @@ When reporting issues, always include:
 * Snapshots or log files if needed
 
 Because the issues are open to the public, when submitting files, be sure to remove any sensitive information, e.g. user name, password, IP address, and company name. You can
-replace those parts with "REDACTED" or other strings like "****".
+replace those parts with "REDACTED" or other strings like "****".

README.md

@@ -54,10 +54,6 @@ Access MaxKB web interface at `http://your_server_ip:8080` with default admin cr
 - LLM Framework：[LangChain](https://www.langchain.com/)
 - Database：[PostgreSQL + pgvector](https://www.postgresql.org/)
 
-## Star History
-
-[![Star History Chart](https://api.star-history.com/svg?repos=1Panel-dev/MaxKB&type=Date)](https://star-history.com/#1Panel-dev/MaxKB&Date)
-
 ## License
 
 Licensed under The GNU General Public License version 3 (GPLv3)  (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

README_CN.md

@@ -4,7 +4,7 @@
     <a href="https://trendshift.io/repositories/9113" target="_blank"><img src="https://trendshift.io/api/badge/repositories/9113" alt="1Panel-dev%2FMaxKB | Trendshift" style="width: 250px; height: auto;" /></a>
 </p>
 <p align="center">
-  <a href="README_EN.md"><img src="https://img.shields.io/badge/English_README-blue" alt="English README"></a>
+  <a href="README.md"><img src="https://img.shields.io/badge/English_README-blue" alt="English README"></a>
   <a href="https://www.gnu.org/licenses/gpl-3.0.html#license-text"><img src="https://img.shields.io/github/license/1Panel-dev/maxkb?color=%231890FF" alt="License: GPL v3"></a>
   <a href="https://github.com/1Panel-dev/maxkb/releases/latest"><img src="https://img.shields.io/github/v/release/1Panel-dev/maxkb" alt="Latest release"></a>
   <a href="https://github.com/1Panel-dev/maxkb"><img src="https://img.shields.io/github/stars/1Panel-dev/maxkb?style=flat-square" alt="Stars"></a>

SECURITY.md

@@ -1,17 +1,15 @@
 # 安全说明
 
-如果您发现安全问题，请直接联系我们：
+如果您发现安全问题，请[提交给我们](https://github.com/1Panel-dev/MaxKB/security/advisories/new)。
 
-- support@fit2cloud.com
-- 400-052-0755
+谢绝只有静态代码分析，未经完整PoC验证的安全建议。
 
 感谢您的支持！
 
 # Security Policy
 
-All security bugs should be reported to the contact as below:
+Any security issue please [submit a vulnerability report](https://github.com/1Panel-dev/MaxKB/security/advisories/new).
 
-- support@fit2cloud.com
-- 400-052-0755
+We do not accept security recommendations based solely on static code analysis without complete PoC validation.
 
-Thanks for your support!
+Thanks for your support!

apps/application/chat_pipeline/step/chat_step/impl/base_chat_step.py

@@ -384,10 +384,11 @@ def _handle_mcp_request(
                 if tool is None or tool.is_active is False:
                     continue
                 executor = ToolExecutor()
+                init_params_default_value = {i["field"]: i.get('default_value') for i in tool.init_field_list}
                 if tool.init_params is not None:
-                    tool_init_params = json.loads(rsa_long_decrypt(tool.init_params))
+                    tool_init_params = init_params_default_value | json.loads(rsa_long_decrypt(tool.init_params))
                 else:
-                    tool_init_params = {i["field"]: i.get("default_value") for i in tool.init_field_list}
+                    tool_init_params = init_params_default_value
                 tool_config = executor.get_tool_mcp_config(tool, tool_init_params)
 
                 mcp_servers_config[str(tool.id)] = tool_config