Skip to content

Commit fdaeda1

Browse files
authored
Update overview docs to fit version 2.1.0 (#2906)
1 parent 03774ff commit fdaeda1

8 files changed

Lines changed: 668 additions & 301 deletions

File tree

Lines changed: 42 additions & 47 deletions
Original file line numberDiff line numberDiff line change
@@ -1,78 +1,73 @@
11
# Key Features
22

3-
Nexent provides powerful capabilities for building and deploying AI agents with minimal effort. Here are the core features that make Nexent unique.
3+
Nexent v2.0 delivers powerful capabilities for building and deploying AI agents. Here are the core features that make Nexent unique.
44

5-
## 🧠 Smart Agent Prompt Generation
5+
## ⚙️ Multi-Model Integration
66

7-
Turn plain language into runnable prompts. Nexent automatically chooses the right tools and plans the best action path for every request.
7+
Nexent is compatible with any OpenAI-compatible model provider, offering one-stop coverage for LLM, Embedding, VLM, STT, and TTS model types. Supports seamless synchronization with the ModelEngine platform, with built-in connection monitoring and automatic failover. The platform supports connecting to any service that follows the OpenAI API protocol, making it easy to diversify models or switch to domestic alternatives.
88

9-
![Feature 1](../../assets/Feature1.png)
9+
## 🤖 Zero-Code Agent Generation
1010

11-
## ⚡ Scalable Data Process Engine
11+
Describe your needs in natural language and Nexent automatically transforms them into executable agent configurations. The system intelligently selects appropriate tools, plans the optimal execution path, and generates professional prompts. No code, no drag-and-drop configuration — experience true "what you imagine is what you get" agent creation. Agents can also be imported and exported for easy sharing and reuse. Built-in debugging provides online testing so you can iterate and refine rapidly.
1212

13-
Process 20+ data formats with fast OCR and table structure extraction, scaling smoothly from a single process to large-batch pipelines.
13+
## 🤝 A2A Protocol & Agent Collaboration
1414

15-
![Feature 2](../../assets/Feature2.png)
15+
Nexent supports the **Agent-to-Agent (A2A)** communication protocol, enabling seamless multi-agent collaboration. A main agent can invoke sub-agents to complete specific tasks; once a sub-agent finishes execution, results are aggregated back to the main agent. Multiple collaborative sub-agents can be configured, each with its own toolset, model configuration, and execution strategy — making it easy to build complex distributed agent workflows.
1616

17-
## 📚 Personal-Grade Knowledge Base
17+
## 🧠 Layered Memory Architecture
1818

19-
Import files in real time, auto-summarise them, and let agents access both personal and global knowledge instantly, also knowing what it can get from each knowledge base.
19+
Intelligent context management is the key to agents that truly understand you. Nexent provides a two-tier memory system:
2020

21-
![Feature 3](../../assets/Feature3.png)
21+
- **User-Level Memory**: Personal preferences, habits, and usage patterns
22+
- **User-Agent Memory**: Collaboration history and context for a specific user with a specific agent
2223

23-
## 🌐 Internet Knowledge Search
24+
The system automatically extracts key information from conversations to generate memory entries — no manual input required. Memory entries can also be added or modified manually for greater flexibility. Smart retrieval ensures every conversation automatically pulls in the most relevant contextual memories, enabling truly personalized service.
2425

25-
Connect to 5+ web search providers so agents can mix fresh internet facts with your private data.
26+
## 📝 Progressive Skill Disclosure
2627

27-
![Feature 4](../../assets/Feature4.png)
28+
Nexent introduces a **Progressive Skill Disclosure** mechanism. As users input tasks, the system dynamically reveals the most relevant Skill suggestions based on the current context — helping users quickly find the tools and methods best suited to the current task. This mechanism enables newcomers to progressively explore system capabilities without adding operational complexity for advanced users.
2829

29-
## 🔍 Knowledge-Level Traceability
30+
## 🗄️ Personal-Grade Knowledge Base
3031

31-
Serve answers with precise citations from web and knowledge-base sources, making every fact verifiable.
32+
Create personal knowledge bases on the Nexent platform. Import files in real time with automatic parsing and vectorization, enabling agents to access private data instantly. Supports 20+ document formats including text, PDF, Word, PowerPoint, Excel, and CSV — with fast OCR and table structure extraction built in. Each knowledge base automatically generates its own summary, helping the agent accurately determine when to retrieve from it. Fine-grained access controls can be set: private, department-wide, or organization-wide visibility.
3233

33-
![Feature 5](../../assets/Feature5.png)
34+
## 🔧 MCP Tool Ecosystem
3435

35-
## 🎭 Multimodal Understanding & Dialogue
36+
Nexent builds its tool ecosystem on the **Model Context Protocol (MCP)** — described as the "USB-C of AI" — a universal interface standard for connecting AI agents to the external world.
3637

37-
Speak, type, files, or show images. Nexent understands voice, text, and pictures, and can even generate new images on demand.
38+
- Add third-party MCP services quickly via URL or JSON configuration
39+
- Develop local MCP tools with LangChain integrations and custom Python plugins
40+
- Hot-swap tools, models, and toolchains without touching core code
41+
- Built-in tool testing lets you verify whether tools work as expected before building an agent
3842

39-
![Feature 6](../../assets/Feature6.png)
43+
## 🌐 Internet Knowledge Integration
4044

41-
## 🔧 MCP Tool Ecosystem
45+
Connect to multiple web search providers so agents can blend the freshest internet information with your private data. Hybrid search mode balances real-time accuracy with relevance.
4246

43-
Drop in or build Python plug-ins that follow the MCP spec; swap models, tools, and chains without touching core code.
47+
## 🔍 Knowledge Traceability & Citations
4448

45-
![Feature 7](../../assets/Feature7.png)
49+
Every answer comes with precise citations from web search results or knowledge base documents, making every fact transparent and verifiable. Source information is fully traceable with one click, building trust in agent responses.
4650

47-
## 🏗️ Architecture Benefits
51+
## 🎭 Multimodal Interaction
4852

49-
### ⚡ Distributed Processing Capabilities
50-
- **Asynchronous Architecture**: High-performance asynchronous processing based on asyncio
51-
- **Multi-threading Safety**: Thread-safe concurrent processing mechanisms
52-
- **Celery Integration**: Optimized for distributed task queues
53-
- **Batch Optimization**: Intelligent batch operations to reduce network overhead
53+
Supports multiple input modes: voice, text, images, and files. Agents can understand voice, text, and images, and can generate new images on demand — delivering a truly natural multimodal conversation experience.
5454

55-
### 🏢 Enterprise-grade Scalability
56-
- **Modular Design**: Loose-coupled module architecture for easy extension
57-
- **Plugin-based Tools**: Standardized tool interfaces for rapid integration
58-
- **Configuration Management**: Flexible configuration system supporting multi-environment deployment
59-
- **Monitoring Friendly**: Comprehensive logging and status monitoring
55+
## 🔢 Agent Version Management
6056

61-
### 🚀 High-performance Optimization
62-
- **Connection Pooling**: Intelligent reuse of database and HTTP connections
63-
- **Memory Management**: Stream processing of large files and memory optimization
64-
- **Concurrency Control**: Intelligent concurrency limiting and load balancing
65-
- **Caching Strategy**: Multi-layer caching to improve response speed
57+
A comprehensive version control system supports agent iteration and historical rollback. Every version is independently archived; view change history, compare versions, and roll back whenever needed. Agent configurations can also be imported and exported in JSON format, enabling seamless migration across environments and smooth team collaboration.
6658

67-
For detailed information about Nexent's software architecture and technical advantages, see our **[Software Architecture](./software-architecture)** guide.
59+
## 🏪 Agent Market
6860

69-
## 🎯 Use Cases
61+
A built-in agent marketplace brings together high-quality agents from both official and community creators. Download with one click to use immediately, or integrate them as sub-agents into your own agent workflows to rapidly build complex applications.
7062

71-
Nexent is designed for various scenarios including:
72-
- **Business Intelligence**: Automated data analysis and reporting
73-
- **Customer Support**: Intelligent chat agents with knowledge base integration
74-
- **Content Processing**: Document analysis, summarization, and extraction
75-
- **Research Assistance**: Academic paper analysis and information synthesis
76-
- **Personal Productivity**: Smart assistants for daily tasks and information management
63+
## 👥 Multi-Tenant RBAC & User Management
7764

78-
For detailed agent scenarios and real-world implementations, see our **[MCP Ecosystem Use Cases](../mcp-ecosystem/use-cases)**.
65+
Nexent provides a complete multi-tenant, role-based permission management system:
66+
67+
- **Four Roles**: Super Administrator, Tenant Administrator, Developer, and Regular User — each with clearly defined responsibilities
68+
- **Multi-Tenant Isolation**: Complete data isolation between tenants, with platform-wide management support
69+
- **User Group Mechanism**: Manage resources and access permissions through groups, supporting flexible permission delegation
70+
- **Invitation Code Mechanism**: Controlled registration safeguards platform security
71+
- **Resource-Level Permissions**: Fine-grained access control on agents, knowledge bases, and more — down to the user group level
72+
73+
For detailed information about Nexent's software architecture and technical advantages, see our **[Software Architecture](./software-architecture)** guide.

doc/docs/en/getting-started/overview.md

Lines changed: 34 additions & 25 deletions
Original file line numberDiff line numberDiff line change
@@ -17,10 +17,10 @@ Nexent is a zero-code platform for auto-generating production-grade AI agents, b
1717

1818
> *If you want to go fast, go alone; if you want to go far, go together.*
1919
20-
We have released **Nexent v1**, and the platform is now relatively stable. However, there may still be some bugs, and we are continuously improving and adding new features. Stay tuned: we will announce **v2.0** soon!
20+
We have released **Nexent v2.0** — a major upgrade over v1.0. This release brings A2A protocol support, progressive Skill disclosure, layered memory architecture, full-featured user management with RBAC, agent version management, and the Agent Market. Core capabilities like knowledge base integration, multimodal interaction, and the MCP tool ecosystem have been significantly enhanced. The platform is maturing rapidly and we welcome your feedback.
2121

22-
* **🗺️ Check our [Feature Map](https://github.com/orgs/ModelEngine-Group/projects/6)** to explore current and upcoming features.
23-
* **🔍 Try the current build** and leave ideas or bugs in the [Issues](https://github.com/ModelEngine-Group/nexent/issues) tab.
22+
- **🗺️ Check our [Feature Map](https://github.com/orgs/ModelEngine-Group/projects/6)** to explore current and upcoming features.
23+
- **🔍 Try the current build** and leave ideas or bugs in the [Issues](https://github.com/ModelEngine-Group/nexent/issues) tab.
2424

2525
> *Rome wasn't built in a day.*
2626
@@ -32,15 +32,21 @@ Most of all, we need visibility. Star ⭐ and watch the [GitHub repository](http
3232

3333
## ✨ Key Features
3434

35-
Nexent offers a comprehensive set of features for building powerful AI agents:
36-
37-
- **🤖 Smart Agent Generation** - Zero-code agent creation using natural language
38-
- **📊 Scalable Data Processing** - Handle 20+ file formats with intelligent extraction
39-
- **🧠 Personal Knowledge Base** - Real-time file import with auto-summarization
40-
- **🌐 Internet Integration** - Connect to multiple search providers and web sources
41-
- **🔍 Knowledge Traceability** - Precise citation and source verification
42-
- **🎭 Multimodal Support** - Voice, text, images, and file processing
43-
- **🔧 MCP Ecosystem** - Extensible tool integration and custom development
35+
Nexent v2.0 delivers a comprehensive feature set for building powerful AI agents:
36+
37+
- **⚙️ Multi-Model Integration** — OpenAI-compatible any provider, with full Embedding/VLM/STT/TTS support
38+
- **🤖 Zero-Code Agent Generation** — Describe in plain language, deploy in one click
39+
- **🤝 A2A Agent Collaboration** — Agent-to-Agent protocol for seamless multi-agent workflows
40+
- **🧠 Layered Memory Architecture** — Two-tier memory system with cross-conversation context accumulation
41+
- **📝 Progressive Skill Disclosure** — Context-aware tool suggestions that reveal as you go
42+
- **🗄️ Personal-Grade Knowledge Base** — 20+ format document import with intelligent retrieval
43+
- **🔧 MCP Tool Ecosystem** — Plug-and-play extensibility with custom tool development
44+
- **🌐 Internet Knowledge Integration** — Multi-source hybrid search blending real-time web with private data
45+
- **🔍 Knowledge-Level Traceability** — Precise citations and verifiable sources on every answer
46+
- **🎭 Multimodal Interaction** — Voice, text, images, and files for fully natural conversations
47+
- **🔢 Agent Version Management** — Version iteration and rollback for safe, controlled deployments
48+
- **🏪 Agent Market** — Official and community agents ready to install and use
49+
- **👥 Multi-Tenant RBAC** — Tenant isolation, role-based permissions, and fine-grained resource access
4450

4551
For detailed feature information and examples, see our **[Features Guide](./features)**.
4652

@@ -49,30 +55,33 @@ For detailed feature information and examples, see our **[Features Guide](./feat
4955
Nexent adopts a modern distributed microservices architecture designed to provide high-performance, scalable AI agent platform. The entire system is based on containerized deployment, supporting cloud-native and enterprise-grade application scenarios.
5056

5157
### 🌐 Layered Architecture Design
52-
- **Frontend Layer** - Modern user interface built with Next.js + React + TypeScript
53-
- **API Gateway Layer** - FastAPI high-performance web framework for request routing and load balancing
54-
- **Business Logic Layer** - Agent management, conversation management, knowledge base management, and model management
55-
- **Data Layer** - Distributed storage architecture with PostgreSQL, Elasticsearch, Redis, and MinIO
58+
59+
- **Frontend Layer** — Modern user interface built with Next.js + React + TypeScript
60+
- **API Gateway Layer** — FastAPI high-performance web framework for request routing and load balancing
61+
- **Business Logic Layer** — Agent management, conversation management, knowledge base management, and model management
62+
- **Data Layer** — Distributed storage architecture with PostgreSQL, Elasticsearch, Redis, and MinIO
5663

5764
### 🚀 Core Service Architecture
58-
- **Agent Services** - Agent generation and execution based on SmolAgents framework
59-
- **Data Processing Services** - Real-time and batch processing supporting 20+ file formats
60-
- **MCP Ecosystem** - Standardized tool interfaces and plugin architecture
65+
66+
- **Agent Services** — Agent generation and execution based on SmolAgents framework
67+
- **Data Processing Services** — Real-time and batch processing supporting 20+ file formats
68+
- **MCP Ecosystem** — Standardized tool interfaces and plugin architecture
6169

6270
### ⚡ Distributed Features
63-
- **Asynchronous Processing** - High-performance async processing architecture based on asyncio
64-
- **Microservices Design** - Service decoupling with independent scaling and deployment
65-
- **Containerized Deployment** - Docker Compose service orchestration supporting cloud-native deployment
71+
72+
- **Asynchronous Processing** — High-performance async processing architecture based on asyncio
73+
- **Microservices Design** — Service decoupling with independent scaling and deployment
74+
- **Containerized Deployment** — Docker Compose service orchestration supporting cloud-native deployment
6675

6776
For detailed architectural design and technical implementation, see our **[Software Architecture](./software-architecture)**.
6877

6978
## ⚡ Quick Start
7079

7180
Ready to get started? Here are your next steps:
7281

73-
1. **📋 [Installation & Deployment](../quick-start/installation)** - System requirements and deployment guide
74-
2. **🔧 [Developer Guide](../developer-guide/overview)** - Build from source and customize
75-
3. **[FAQ](../quick-start/faq)** - Common questions and troubleshooting
82+
1. **📋 [Installation & Deployment](../quick-start/installation)** System requirements and deployment guide
83+
2. **🔧 [Developer Guide](../developer-guide/overview)** Build from source and customize
84+
3. **[FAQ](../quick-start/faq)** Common questions and troubleshooting
7685

7786
## 💬 Community & contact
7887

0 commit comments

Comments
 (0)