Dmmsys Foundation AI Agent
by Dmmsys guoyihot@outlook.com 124358678@qq.com http://www.dmmsys.com$ 96.95
| Availability |
Odoo Online
Odoo.sh
On Premise
|
| Odoo Apps Dependencies |
Discuss (mail)
|
| Lines of code | 3790 |
| Technical Name |
dmmsys_foundation_ai_agent |
| License | OPL-1 |
| Website | http://www.dmmsys.com |
| Versions | 17.0 18.0 19.0 |
Dmmsys Foundation AI Agent
AI assistant for Odoo Community Edition. Build a knowledge base from your documents and web pages. Let users ask questions and get answers from your knowledge base. Auto-run routine tasks with simple commands. No Enterprise license required.
Dmmsys Foundation AI Agent — 轻量级 AI 助手
让员工用聊天的方式操作业务数据。上传文件或网页，提问就能得到答案。配置 AI 助手，帮您查数据、改单据、建订单。用简单指令自动执行日常任务。无需企业版许可。
What You Can Do
Everyday tasks, made simple with AIAsk Questions, Get Answers
Talk to your business data like talking to a colleagueChat with Your Data
"What was last month's total sales?" "Show me overdue invoices." Get instant answers without running reports.
Search Files & Web Pages
Upload product manuals, policies, or any documents. Link web pages. Ask questions and AI finds answers from your knowledge base.
Let AI Do the Work
"Create a new customer ABC Company." "Update order SO123 to priority high." AI can read, update, and create records for you.
Automate Routine Tasks
Turn multi-step processes into one-click automationMulti-Step Automation
Set up a sequence of actions — search, read, update, create. AI executes them in order, automatically retrying if something fails.
Custom AI Helpers
Create different AI assistants for different jobs — a finance assistant, a customer support assistant, an operations assistant. Each with its own knowledge and tools.
Use Anywhere
Floating button on every screen. Click, type your request, get help instantly. Works in forms, lists, and chatter.
能帮你做什么
日常工作，交给 AI提问就有答案
像跟同事聊天一样，跟数据对话跟数据聊天
"上个月销售额多少？""哪些订单逾期了？"不用跑报表，直接问就有答案。
搜索文件与网页
上传产品手册、政策文件、操作规范，或链接网页。AI 从知识库里帮你找答案。
让 AI 帮你干活
"新建一个客户 ABC 公司""把订单 SO123 改成紧急优先"AI 帮你查、帮你改、帮你建。
自动执行日常任务
多步流程，一键自动完成多步骤自动化
设置一串动作——查、读、改、建。AI 按顺序执行，失败了自动重试。
定制 AI 助手
为不同岗位创建专属 AI——财务助手、客服助手、运营助手。每个都有自己的知识库和工具。
随时随地使用
每个页面都有浮动按钮。点击、输入、获得帮助。表单、列表、聊天窗口都能用。
For IT Managers
Technical capabilities & requirementsTechnical Capabilities
What's under the hoodAI Agent Core
Configurable agents with topic/role system. Response style control (analytical/balanced/creative). Knowledge sources (files + web pages). Restrict answers to knowledge sources only.
RAG & Vector Search
PostgreSQL vector extension with efficient indexing. Automatic text chunking and embedding generation. Scheduled background processing for batch embedding.
Tool System
Built-in actions for searching, reading, updating, and creating records. Any system action can be turned into an AI tool with a simple flag. Supports structured parameters with validation. Automatic conversion to OpenAI function‑calling format.
Task Orchestration
Define multi‑step workflows with dependency rules. AI executes steps in the correct order and automatically retries on failure. Task templates and execution logs provided.
Integration & Requirements
Deployment and compatibilityLLM & UI
Works with any LLM that supports the OpenAI API (DeepSeek, Qwen, Zhipu, OpenAI, etc.). Mock client for offline testing. Floating action button plus a VS Code style chat window. Model and agent selector. Session persists after page refresh.
Business Integration
Seamlessly integrated into all business models. Customizable entry points for different UI contexts. Preset shortcut prompt buttons for quick access.
Requirements
Designed for Odoo Community Edition as an alternative to Enterprise AI capabilities. PostgreSQL with vector extension recommended for knowledge‑based features. Python requests library required.
给 IT 管理人员
技术能力与环境要求技术能力
核心能力一览AI Agent 核心
可配置的代理（主题+角色系统）。响应风格控制（分析型/平衡型/创造型）。知识源（文件+网页）。可限制仅从知识源回答。
RAG 与向量检索
PostgreSQL 向量扩展 + 高效索引。自动分块与向量化。定时任务批量生成向量。
工具系统
内置查、读、改、建通用动作。任何系统动作都可一键注册为 AI 工具。支持结构化参数与校验。自动转换为 OpenAI 函数调用格式。
任务编排
定义多步骤工作流并设置依赖关系。AI 按正确顺序执行，失败时自动重试。提供任务模板和执行日志。
集成与环境要求
部署与兼容性LLM 与 UI
兼容 OpenAI 接口的 LLM（DeepSeek、通义千问、智谱、OpenAI 等）。支持离线测试的模拟客户端。浮动按钮 + VS Code 风格聊天窗口。模型与代理选择器。刷新后会话持久化。
业务集成
无缝集成所有业务模型。支持多入口的自定义触发规则。预设快捷提示按钮。
环境要求
为 Odoo 社区版设计，是企业版 AI 功能的替代方案。知识库功能建议使用 PostgreSQL + 向量扩展。需要 Python requests 库。
AI Agent Base Capability Comparison
Capabilities across Foundation / Foundation MCP / Professional / Enterprise editions
Module mapping by edition:
• Foundation — dmmsys_foundation_ai_agent: Basic AI agent capabilities (chat / tool / task / RAG / LLM)
• Foundation MCP — dmmsys_foundation_mcp_agent: Foundation AI capabilities plus MCP gateway for external AI integration
• Professional — dmmsys_professional_mcp_agent: Full professional-grade AI platform with audit, evaluation, workflow integration, alerts
• Enterprise — dmmsys_enterprise_ai_agent + dmmsys_enterprise_essence series: Enterprise AI with ontology, self-evolution, budget control, monitoring
| Capability | Foundation | Foundation MCP | Professional | Enterprise |
|---|---|---|---|---|
| Layer 0 — Intent Core (Foundation + MCP Gateway) | ||||
| Agent Core (chat / tool / task) | ✓ | ✓ | ✓ | ✓ |
| LLM Integration | ✓ | ✓ | ✓ | ✓ |
| RAG Knowledge Base | ✓ | ✓ | ✓ | ✓ |
| Prompt Buttons / Composer | ✓ | ✓ | ✓ | ✓ |
| Cognitive Orchestration (Intent Recognition) | ✓ | ✓ | ✓ | ✓ |
| MCP Gateway (External AI Integration) | ✗ | ✓ | ✓ | ✓ |
| Layer 1 — Governance Agent (Professional – Audit, Evaluation, Workflow, Alerts) | ||||
| Audit Log | ✗ | ✗ | ✓ | ✓ |
| Evaluation Suite (Regression Testing) | ✗ | ✗ | ✓ | ✓ |
| Process Integration (Workflow) | ✗ | ✗ | ✓ | ✓ |
| Alert Rules | ✗ | ✗ | ✓ | ✓ |
| Layer 2 — Essence AI (Enterprise – Ontology, Self-Evolution, Budget, Ticket, Monitoring) | ||||
| Heartbeat Health Monitoring | ✗ | ✗ | ✗ | ✓ |
| Budget Control | ✗ | ✗ | ✗ | ✓ |
| Ticket (Human Approval) | ✗ | ✗ | ✗ | ✓ |
| Skill Training Loop (Self-Evolution) | ✗ | ✗ | ✗ | ✓ |
| Business Ontology (Semantic Definition) | ✗ | ✗ | ✗ | ✓ (essence series) |
| Agent Delegation Log | ✗ | ✗ | ✗ | ✓ |
| Monitor Dashboard | ✗ | ✗ | ✗ | ✓ |
| Tool Scan / Call Log | ✗ | ✗ | ✗ | ✓ |
AI Agent 底座能力对照表
Foundation / Foundation MCP / Professional / Enterprise 四层底座能力对比
各层级对应的模块名称：
• Foundation — dmmsys_foundation_ai_agent：基础 AI Agent 核心能力（chat / tool / task / RAG / LLM）
• Foundation MCP — dmmsys_foundation_mcp_agent：基础 AI 能力 + MCP 网关（外部 AI 系统接入）
• Professional — dmmsys_professional_mcp_agent：专业级 AI 平台，增加审计、评估、流程集成、告警等功能
• Enterprise — dmmsys_enterprise_ai_agent + dmmsys_enterprise_essence 系列：企业级 AI，具备本体语义、自进化、预算控制、监控等能力
| 能力维度 | Foundation | Foundation MCP | Professional | Enterprise |
|---|---|---|---|---|
| 第 0 层 — Intent Core (意图驱动核心) – 基础 AI 能力 | ||||
| Agent 核心（chat/tool/task） | ✓ | ✓ | ✓ | ✓ |
| LLM 接入 | ✓ | ✓ | ✓ | ✓ |
| RAG 知识库 | ✓ | ✓ | ✓ | ✓ |
| 提示词按钮 / Composer | ✓ | ✓ | ✓ | ✓ |
| 认知编排（意图识别） | ✓ | ✓ | ✓ | ✓ |
| MCP 网关（外部 AI 接入） | ✗ | ✓ | ✓ | ✓ |
| 第 1 层 —Governance Agent (治理增强版) – 加入审计、权限、审批等治理 | ||||
| 审计日志 | ✗ | ✗ | ✓ | ✓ |
| 评估套件（回归测试） | ✗ | ✗ | ✓ | ✓ |
| 流程集成（Workflow） | ✗ | ✗ | ✓ | ✓ |
| 告警规则 | ✗ | ✗ | ✓ | ✓ |
| 第 2 层 —Essence AI (企业本体版) – 加入本体、PCF、知识图谱、自进化 | ||||
| Heartbeat 健康监控 | ✗ | ✗ | ✗ | ✓ |
| 预算控制 | ✗ | ✗ | ✗ | ✓ |
| 工单（人工审批） | ✗ | ✗ | ✗ | ✓ |
| Skill 训练闭环（自进化） | ✗ | ✗ | ✗ | ✓ |
| 业务本体（语义定义） | ✗ | ✗ | ✗ | ✓（essence 系列） |
| Agent 委派日志 | ✗ | ✗ | ✗ | ✓ |
| 监控仪表板 | ✗ | ✗ | ✗ | ✓ |
| 工具扫描 / 调用日志 | ✗ | ✗ | ✗ | ✓ |
Zero-Token Prompt Buttons
Foundation Edition ships with 4 built-in prompt buttons in the FAB Chat window:
• Summarize this record -LLM — LLM required
• Analyze this record -LLM — LLM required
• Suggest next actions based on the current record state — 0-Token (direct tool execution, no LLM)
• Explain this record -LLM — LLM required
Professional Edition includes additional business-specific prompt buttons (e.g., Return Order Assistant, Purchase Order Assistant, Production Order Assistant, Invoice Dispute Analysis, Payment Suggestion) tailored to each business model. These can also be configured as 0-Token via the same Direct Action Tool mechanism.
Beyond these built-in defaults, the module provides the Direct Action Tool mechanism
on ai.prompt.button: administrators can assign any ai.generic.tool record
to any prompt button. Clicking that button executes the backend method directly,
bypassing the LLM entirely and consuming zero tokens.
This enables deterministic business operations (e.g., checking inventory, closing a period,
releasing a work order) to run instantly without any AI cost.
Prompt buttons without a direct tool assigned continue to use the LLM as normal.
零Token 提示词按钮
Foundation 标准版在 FAB Chat 窗口中内置了 4 个通用提示词按钮，其中：
• Summarize this record -LLM（总结） — 需LLM
• Analyze this record -LLM（分析） — 需LLM
• Suggest next actions based on the current record state（建议下一步操作） — 0-Token，不调LLM
• Explain this record -LLM（解释） — 需LLM
Professional 专业版额外包含针对各业务对象的专用提示词按钮（如退货助手、采购助手、生产工单助手、发票争议分析、付款建议等），
同样可通过 Direct Action Tool 机制配置为 0-Token。
除内置默认配置外，本模块提供了 Direct Action Tool（直接执行工具）机制：
管理员可以在 ai.prompt.button 上为任意提示词按钮绑定一个 ai.generic.tool 记录。
绑定后点击该按钮将直接调用后端方法执行，完全绕过 LLM，消耗 零个 Token。
这使得确定性业务操作（如检查库存、关账、释放工单等）可以即时运行，无任何 AI 成本。
未绑定直接工具的提示词按钮则继续正常使用 LLM。
Professional Business Agents
专业业务智能体
Ready‑to‑use AI workers for core manufacturing operations开箱即用的制造业核心业务 AI 数字员工
| Agent Module 模块名称 |
Functionality 功能说明 |
Default Depends 默认底座依赖 |
|---|---|---|
| dmmsys_mrp_intelligent_scheduling APS 智能排程 | 有限产能排程（OR-Tools CP-SAT 多目标优化）、帕累托方案对比（甘特图）、动态重排、派工单 Finite capacity scheduling (OR‑Tools multi‑objective), Pareto comparison (Gantt), dynamic rescheduling, dispatch orders |
dmmsys_foundation_ai_agent |
| dmmsys_mrp_decision_agent 生产排产 AI 决策 | 物料可用性/产能检查、重排建议、批量下达审批 Material availability & capacity checks, rescheduling suggestions, batch release approval |
dmmsys_professional_mcp_agent |
| dmmsys_monthly_closing_agent 财务月结 AI 数字员工 | 自动化月结流程（采购/销售/库存/生产/总账/资产六大模块的顺序检查、自动修复、期间锁定） Automated month-end closing: sequential check, auto‑fix, period lock for purchasing, sales, inventory, production, general ledger, fixed assets |
dmmsys_enterprise_ai_agent |
| dmmsys_professional_change_agent PLM 工程变更数字员工 | 变更管理四层（问题报告/变更申请/变更通知/变更任务）的 AI 检查/修复/关闭 Four‑layer change management (Issue → Request → Notice → Task) AI inspection, fixing, and closure |
dmmsys_enterprise_ai_agent |
| dmmsys_stock_replenishment_agent 自动补货 AI 决策 | 库存消耗分析、再订货点优化、自动创建采购订单 Consumption pattern analysis, reorder point optimization, automatic purchase order creation |
dmmsys_enterprise_ai_agent |
| dmmsys_stock_reservation_agent 物料预留 AI 决策 | 库存预留、缺料处理、替代料建议 Stock reservation, shortage handling, substitute material suggestions |
dmmsys_enterprise_ai_agent |
Note / 说明： All agents share identical _ai_xxx business logic. The depends only determines which AI infrastructure (audit, evolution, ontology, etc.) the agent can leverage.
所有专业 Agent 的业务逻辑代码完全相同，depends 仅决定该 Agent 可以使用哪些底层 AI 基础设施（审计、自进化、本体推理等）。
Professional Business Agents
Ready‑to‑use AI workers for core manufacturing operations| Agent Module | Functionality | Default Depends |
|---|---|---|
| dmmsys_mrp_intelligent_scheduling APS intelligent scheduling |
Finite capacity scheduling (OR‑Tools CP‑SAT multi‑objective), Pareto comparison (Gantt), dynamic rescheduling, dispatch orders | dmmsys_foundation_ai_agent |
| dmmsys_mrp_decision_agent Production scheduling AI decision |
Material availability & capacity checks, rescheduling suggestions, batch release approval | dmmsys_professional_mcp_agent |
| dmmsys_monthly_closing_agent Financial month‑end closing AI agent |
Automated month‑end closing: sequential check, auto‑fix, period lock for purchasing, sales, inventory, production, general ledger, fixed assets | dmmsys_enterprise_ai_agent |
| dmmsys_professional_change_agent PLM engineering change AI agent |
Four‑layer change management (Issue → Request → Notice → Task) AI inspection, fixing, and closure | dmmsys_enterprise_ai_agent |
| dmmsys_stock_replenishment_agent Automatic replenishment AI decision |
Consumption pattern analysis, reorder point optimization, automatic purchase order creation | dmmsys_enterprise_ai_agent |
| dmmsys_stock_reservation_agent Material reservation AI decision |
Stock reservation, shortage handling, substitute material suggestions | dmmsys_enterprise_ai_agent |
Note: All agents share identical _ai_xxx business logic. The depends only determines which AI infrastructure (audit, evolution, ontology, etc.) the agent can leverage.
Three‑Layer Base Adaptation
All professional agents can run on any base layer. Choose the right base according to your needs.
Foundation
dmmsys_foundation_ai_agent
- Agent Core + LLM + RAG
- Composer + cognitive orchestration
- Basic tool system
Basic AI capabilities
Foundation MCP
dmmsys_foundation_mcp_agent
- All Foundation capabilities
- + MCP Gateway (external AI access)
- Token authentication + IP whitelist
Foundation + MCP gateway
Professional
dmmsys_professional_mcp_agent
- All Foundation MCP capabilities
- + Full audit logs
- + Evaluation suite & regression testing
- + Workflow integration & alert rules
Professional AI platform
Enterprise
dmmsys_enterprise_ai_agent + essence series
- All Professional capabilities
- + Business ontology (
ontology.object/link/action) - + Self‑evolution (nightly pipeline)
- + Budget control & ticket approval
- + Knowledge graph & GraphRAG
Enterprise AI decision foundation
How to Switch the Base Layer
- Locate the agent module’s manifest file (
__manifest__.py). - Find the
dependslist containing the default base module name (e.g.,dmmsys_foundation_ai_agent). - Replace it with the base module name you have purchased (see cards above or the capability comparison table).
- No business code changes required (
_ai_xxxlogic is identical).
Example: To upgrade dmmsys_mrp_intelligent_scheduling to run on Professional base, change depends from 'dmmsys_foundation_ai_agent' to 'dmmsys_professional_mcp_agent'.
Important Notes
- All professional agents share identical business logic (
_ai_xxxmethods). The base layer only adds extra infrastructure capabilities. - You do not need to modify any business code when switching base layers.
- The default dependency in our distribution is always the lowest compatible base (Foundation for scheduling agents, Professional/Enterprise for others). This ensures out‑of‑the‑box functionality for all customers.
- Once you purchase a higher‑level base, simply update the
dependsin the agent’s manifest and the agent will automatically gain the advanced capabilities (audit, evolution, ontology, etc.).
专业业务智能体
开箱即用的制造业核心业务 AI 数字员工| 智能体模块 | 功能说明 | 默认底座依赖 |
|---|---|---|
| dmmsys_mrp_intelligent_scheduling APS 智能排程 |
有限产能排程（OR-Tools CP-SAT 多目标优化）、帕累托方案对比（甘特图）、动态重排、派工单 | dmmsys_foundation_ai_agent |
| dmmsys_mrp_decision_agent 生产排产 AI 决策 |
物料可用性/产能检查、重排建议、批量下达审批 | dmmsys_professional_mcp_agent |
| dmmsys_monthly_closing_agent 财务月结 AI 数字员工 |
自动化月结流程（采购/销售/库存/生产/总账/资产六大模块的顺序检查、自动修复、期间锁定） | dmmsys_enterprise_ai_agent |
| dmmsys_professional_change_agent PLM 工程变更数字员工 |
变更管理四层（问题报告/变更申请/变更通知/变更任务）的 AI 检查/修复/关闭 | dmmsys_enterprise_ai_agent |
| dmmsys_stock_replenishment_agent 自动补货 AI 决策 |
库存消耗分析、再订货点优化、自动创建采购订单 | dmmsys_enterprise_ai_agent |
| dmmsys_stock_reservation_agent 物料预留 AI 决策 |
库存预留、缺料处理、替代料建议 | dmmsys_enterprise_ai_agent |
说明： 所有专业 Agent 的业务逻辑代码（_ai_xxx）完全相同，depends 仅决定该 Agent 可以使用哪些底层 AI 基础设施（审计、自进化、本体推理等）。
三层底座适配方式
所有专业业务智能体均可在任意底座层上运行。根据需求选择最合适的底座。
Foundation
dmmsys_foundation_ai_agent
- Agent 核心 + LLM + RAG
- Composer + 认知编排
- 基础工具系统
基础 AI 能力
Foundation MCP
dmmsys_foundation_mcp_agent
- 包含 Foundation 全部能力
- + MCP 网关（外部 AI 接入）
- Token 认证 + IP 白名单
基础能力 + MCP 网关
Professional
dmmsys_professional_mcp_agent
- 包含 Foundation MCP 全部能力
- + 完整审计日志
- + 评估套件 & 回归测试
- + 流程集成 & 告警规则
专业级 AI 平台
Enterprise
dmmsys_enterprise_ai_agent + essence 系列
- 包含 Professional 全部能力
- + 业务本体 (
ontology.object/link/action) - + 自进化引擎（夜间流水线）
- + 预算控制 & 工单审批
- + 知识图谱 & GraphRAG
企业级 AI 决策底座
如何切换底座
- 找到专业 Agent 模块的 manifest 文件（
__manifest__.py）。 - 找到
depends列表，它当前包含默认底座模块名（例如dmmsys_foundation_ai_agent）。 - 将其替换为您购买的底座模块名（参考上方卡片或底座能力对照表）。
- 无需修改任何业务代码（
_ai_xxx逻辑完全相同）。
示例： 将 dmmsys_mrp_intelligent_scheduling 升级到 Professional 底座，只需将 depends 中的 'dmmsys_foundation_ai_agent' 改为 'dmmsys_professional_mcp_agent'。
重要说明
- 所有专业业务智能体的业务逻辑（
_ai_xxx方法）完全相同，底座仅提供额外的基础设施能力。 - 切换底座时，无需修改任何业务代码。
- 我们发布时默认使用最低兼容底座（排程 Agent 用 Foundation，其他用 Professional/Enterprise），确保所有客户开箱即用。
- 当您购买了更高级的底座后，只需修改 manifest 中的
depends，该 Agent 就会自动获得审计、自进化、本体等高级能力。
Version History / 版本历史
| Version | Date | Changes |
|---|---|---|
| 1.0 | 2026-06-07 |
First release. FAB Chat with 4 built-in prompt buttons (3 LLM + 1 0-Token). 首次发布。FAB Chat 内置 4 个提示词按钮（3 个需 LLM + 1 个 0-Token）。 |
Screenshots
No. Due to Odoo front-end changes in each version, front-end modules do not support cross-version use.
Yes. The module is designed for Odoo Community Edition as an alternative to Enterprise AI features.
PostgreSQL with vector extension is recommended for RAG features. Python requests library is required.
Product and Services
| Availability |
Odoo Online
Odoo.sh
On Premise
|
| Odoo Apps Dependencies |
Discuss (mail)
|
| Lines of code | 3790 |
| Technical Name |
dmmsys_foundation_ai_agent |
| License | OPL-1 |
| Website | http://www.dmmsys.com |
Odoo Proprietary License v1.0 This software and associated files (the "Software") may only be used (executed, modified, executed after modifications) if you have purchased a valid license from the authors, typically via Odoo Apps, or if you have received a written agreement from the authors of the Software (see the COPYRIGHT file). You may develop Odoo modules that use the Software as a library (typically by depending on it, importing it and using its resources), but without copying any source code or material from the Software. You may distribute those modules under the license of your choice, provided that this license is compatible with the terms of the Odoo Proprietary License (For example: LGPL, MIT, or proprietary licenses similar to this one). It is forbidden to publish, distribute, sublicense, or sell copies of the Software or modified copies of the Software. The above copyright notice and this permission notice must be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
Please log in to comment on this module