Article body
Full article
A new generation of Agentic BI from HENGSHI Technology.
1. What Is HENGSHI CLI?
HENGSHI CLI, exposed externally through the unified command identifier hbi, is HENGSHI’s command-line execution system for the AI era. It consolidates the full range of BI operations, including data access, data queries, dashboard creation, access control, and system operations, into a standardized command system. It also ships with repository-hosted skill packages designed for AI Agents.
With HENGSHI CLI, AI Agents and cloud-based agents are no longer limited to guidance or advisory responses. They can complete the full data-analysis workflow directly in the terminal, connecting data questions to real BI delivery.
2. Core CLI Capabilities
Through the unified hbi entry point, the command system covers the full BI workflow: data-source connection, data management, data analysis, dashboard creation, visualization configuration, user permissions, and daily operations. Every business action can be executed through standardized commands, forming a connected closed-loop work path.
The tool also includes repository-native skill packages. Operating rules, execution logic, and scenario workflows are captured as reusable execution plans. AI does not need to repeatedly rediscover rules or infer procedures. It can invoke standard skills directly, reducing the cost of building automated workflows and lowering the chance of trial-and-error failures.
Engineering-Grade Security Protection
HENGSHI CLI is designed with enterprise data-security requirements in mind. Sensitive credentials such as accounts and keys are stored in system credential-management tools, with support for mainstream single sign-on and authorization methods. This prevents sensitive information from being exposed at the source.
For high-risk operations such as permission changes, feature deployment, and environment migration, the system provides dry-run capability. Changes are simulated first and can be reviewed by a person before they take effect. Every operation remains traceable and auditable.
Standardized Result Output
All operation results are returned in machine-friendly formats with clear and consistent structure. AI does not need to spend extra effort interpreting non-standard output. Data reading, information extraction, and cross-system flow become more efficient, and results can connect smoothly with automation scripts and business systems.
Real-Time Human-AI Collaboration
Configuration and content changes made in the terminal can be synchronized in real time to the visual interface. Managers and business users can immediately view what the AI has produced and verify details in the front-end interface. This creates an efficient collaboration model in which AI executes and people review.
3. CLI Skill Packages
HENGSHI CLI is not only a standalone tool. It is delivered with a layered skill system that captures industry practice, operational boundaries, and execution requirements. The skills follow the principle of reading the rules before executing actions, preventing AI from guessing how to operate. Capabilities are divided by business domain rather than piled into lengthy instructions, allowing different AI systems to find the right workflow for each task.
The skill system has four layers that cover BI scenarios end to end.
Foundation Layer
This layer handles identity authentication, global settings, workspace management, accounts, and organizational permissions. It helps AI understand the platform’s basic rules and the relationships among permissions, spaces, and accounts before execution begins.
Data and Semantics Layer
This layer covers data-source connection, dataset management, data modeling, business-metric definition, and data query analysis. It focuses on the full data-processing chain and supports AI in data access, computation, and verification.
Visualization Delivery Layer
This layer focuses on dashboard planning and construction, page layout optimization, data-alert configuration, and operations for intelligent data assistants. It supports the design, delivery, and ongoing management of visual reports.
Automation Orchestration Layer
This layer connects data pipelines, online analysis documents, scheduled tasks, and complex cross-module workflows. It enables AI to orchestrate multi-step, cross-domain, long-running automation for diverse BI operations and delivery needs.
4. Standard CLI Workflow
By combining commands with skill rules, HENGSHI CLI lets AI complete the full workflow from environment discovery to final delivery in a standardized, independent, and auditable way.
Discover the Current State and Business Context
AI first inventories existing platform resources, including connected data sources, datasets, application projects, report assets, and historical data. This gives it a complete view of the environment and business requirements before it acts.
Build Assets and Produce Deliverables
Based on the actual business need, AI then completes data configuration, data analysis and calculation, dashboard creation, page layout, and chart design. It turns an analytics requirement into a usable BI deliverable.
Apply Security Controls and Complete Delivery
After the deliverable is created, AI configures access permissions and user groups as needed. High-risk operations are first validated through dry runs. Once confirmed, they can enter the automation workflow and go live with full operation records, protecting business stability.
5. Adapting to Current AI Runtime Forms and the Future
HENGSHI CLI is built on general terminal-tool invocation. It is not tied to a single AI product. It can adapt to today’s mainstream Agents while remaining extensible for future AI runtime forms.
Coding Agents
Coding Agents work in local development environments, code repositories, and automated release flows. They are good at workflow construction, script development, and system integration. With CLI, they can connect data preparation, report planning, and permission configuration into standardized processes, integrating deeply with project delivery and version iteration. This is especially useful for project-based BI engineering implementation.
Resident Agents
Resident Agents are deployed in the cloud, collaboration platforms, or backend services. They retain long-running context and continuously receive business requests. These Agents can use CLI as a daily work tool for routine data analysis, temporary report creation, periodic operations, and always-on automation services.
Future Compatibility
AI product forms will keep changing, but tool invocation and terminal interaction will remain foundational Agent capabilities. HENGSHI CLI’s standardized interface, skill system, and human-AI collaboration model do not depend on a specific AI framework. Future coding Agents, cloud Agents, and distributed workflow orchestration engines can connect to this system as long as they support basic tool invocation. This protects the long-term stability of enterprise AI+BI architecture.
Conclusion
HENGSHI CLI reconstructs the collaboration model between AI and BI platforms. It is not a simple conversational tool, but an engineering-grade execution layer that connects people, AI, and full business workflows.
Through standardized commands, layered skills, security mechanisms, and real-time collaboration, different AI Agents can independently complete BI engineering work. Repetitive configuration, construction, and operations tasks can move into automation.
Because its foundation is general and extensible, HENGSHI CLI can support today’s AI products and adapt to future intelligent technologies, becoming a stable execution base for the integration of AI and BI in the enterprise.