Industry First: HENGSHI CLI Is Released, Opening the Era of Agentic BI Autopilot

As AI agents evolve at high speed, enterprise business intelligence is standing at a historic crossroads. Mainstream BI vendors have already equipped their products with ChatBI-style intelligent Q&A, greatly improving flexibility for business users. But as business complexity increases and automation requirements rise, traditional BI still faces a fundamental problem: even highly capable large models cannot independently and smoothly complete end-to-end data analysis tasks through traditional BI interfaces.

The industry’s common answer has been to let agents call complex REST APIs. That approach quickly exposes fatal bottlenecks: APIs are highly fragmented, response formats are inconsistent, and agents must painfully handle authentication, rate limits, and retry logic. At the same time, any API change requires code changes, making systems fragile and hard to maintain.

As AI agents enter the core workflow of enterprise software, every analytics platform must answer a new question:

If the future executor of analysis tasks is not only a person but also an AI agent, how should the analytics platform be operated directly?

Based on these insights, we officially release HENGSHI CLI, code-named babayaga, the industry’s first BI command-line tool designed specifically for AI agents.

This enterprise-grade command-line tool is built in Rust. It is not a simple feature add-on, but a foundational reconstruction of product automation under the HENGSHI NEXT strategy. With a focus on unified structure, efficiency, security, and extensibility, it breaks the interaction boundary of traditional BI tools, moves data operations from manual triggers to automated flow, injects new momentum into enterprise digital transformation, and redefines how BI tools connect with business and technology.

From Human-Machine Interaction to Machine-Machine Interaction (Agent-First)

HENGSHI CLI overturns the traditional API wrapping logic. Everything is designed to pave the way for agents.

Unified structure, zero parsing cost: Instead of forcing agents to find different endpoints for different objects, the CLI uses one unified command structure to operate all business objects. It supports highly structured JSON, YAML, and Table output formats, reducing the agent’s parsing cost to nearly zero.

SSE real-time synchronization, what you see is what you get: This is a major architectural breakthrough. When an agent modifies a dashboard or chart through the CLI at the system layer, the backend pushes changes to the front end through Server-Sent Events (SSE). The Web UI refreshes instantly without manual action, letting business users immediately see the agent’s work.

Extreme performance: The CLI is rebuilt in Rust and runs as a single binary with zero external dependency. Startup time is under 10 milliseconds, and most commands complete within 100 milliseconds, providing a strong foundation for high-concurrency agent task scheduling.

With a minimalist architecture, real-time collaboration, and extreme performance, the CLI gives enterprise AI agents an efficient, reliable, and seamless automation execution base.

Security and Full-Scenario Coverage

As a high-speed data engine for medium and large enterprises, we are opening capabilities while establishing the highest standard of security and object control.

Absolute credential security: HENGSHI CLI eliminates the risky practice of writing tokens into plaintext configuration files. Authentication credentials are stored directly in the operating system’s system-level keyring, such as macOS Keychain, Linux libsecret, or Windows Credential Manager. Credentials do not land on disk, and the design fully supports OAuth and enterprise SSO systems.

Full coverage of BI objects: The CLI is not limited to simple queries. It has complete control over HENGSHI SENSE core assets. An AI agent can complete the following operations with a single command:

• Data connection: connection create rapidly connects MySQL and other data sources.
• Semantic construction: dataset create and measure create automatically define datasets and business metrics.
• Visualization rendering: dashboard create and chart create automatically generate dashboards and chart components with a refined technology aesthetic.

From data access to semantic modeling and visualization, the entire chain is secure and controllable, allowing enterprise AI and automated analytics to land efficiently and safely at every step.

Fully Freeing BI Engineering Implementation

The core of digital transformation is using technology to serve business and letting data create value. The release of HENGSHI CLI marks the extension of BI usage scenarios from front-end clicks to backend automation, matching the most advanced workflows.

Fully automated report generation through SSE: With the innovative Server-Sent Events (SSE) real-time synchronization mechanism, when a person enters “generate a monthly sales analysis” in a chat window, the agent behind the scenes can call the CLI seamlessly and complete connection creation, dataset creation, and chart creation within seconds. The front-end page then refreshes automatically and displays the result.

CI/CD and agile BI operations: With the CLI, cross-environment migration and batch export backup of BI assets become as simple as code deployment. Operations teams can write dashboard updates directly into GitHub Actions and other automation pipelines.

Automated linkage with monitoring and alerts: When an external monitoring script detects an anomaly, such as a CPU spike or business circuit break, it can trigger CLI commands to update relevant charts. The front end broadcasts changes in real time so every stakeholder gets situational awareness immediately.

Flexible output formats: The CLI supports JSON, YAML, and Table output. JSON is especially suitable for zero-parsing-cost structured processing by AI agents. YAML is convenient for configuration management, while Table provides human-friendly readability.

CLI brings data automation and intelligence into practice, giving enterprises an efficient and flexible digital-intelligence foundation for transformation.

Not Only for AI: Reshaping How Analytics Teams Work

Although HENGSHI CLI is positioned around Agent First, its value is not limited to AI agents. It will also significantly change the way development teams, analytics teams, and platform administrators work.

For example, many tasks that previously relied on repeated manual clicking can now become more efficient and standardized:

• Rapidly initialize data connections and lower the barrier to data access
• Automatically create analytics datasets and reduce manual configuration
• Generate dashboards in batches and improve the output efficiency of analytics assets
• Migrate analytics assets across environments and maintain environment consistency
• Manage objects through scripts and improve traceability
• Connect the analytics platform seamlessly to CI/CD and automation workflows

This means HENGSHI SENSE is evolving from an analytics tool into analytics platform infrastructure that can be engineered, automated, and orchestrated. This is especially important for large enterprises. A truly mature data analytics platform of the future must do more than draw charts and analyze data. It must also provide:

• Automation capability: keeping the full process efficient with less manual intervention
• Platform governance capability: ensuring compliance, stability, and manageability
• Engineering integration capability: connecting seamlessly with the enterprise technology stack
• AI collaboration capability: linking with AI agents to release the potential of intelligent analytics

The CLI is a key entry point to this layer of capability.

HENGSHI SENSE Has Entered the Next Stage of AI Native Analytics

HENGSHI CLI is now officially released, and one trend is becoming increasingly clear:

Enterprise software in the future must serve not only people, but also AI.

Future analytics platforms should not only provide interfaces for people to view and operate. They must also provide an operational interface layer for AI. In this sense, the release of HENGSHI CLI is not just a new feature launch. It represents Hengshi’s redefinition of the platform boundary of HENGSHI SENSE:

after evolving from a BI tool into an intelligent analytics platform, it is now moving further toward an AI Native analytics platform.

This is not only an upgrade in interface form, but also a ticket to the Agentic BI era. When powerful instruction sets are placed in the hands of AI agent teams, the implementation model of BI will be rewritten. Welcome to the autopilot era of BI.

If you would like to learn more about HENGSHI CLI capabilities, typical scenarios, or partnership opportunities, please contact us. Together, we can explore a new data analytics paradigm for the AI agent era.

Recommended Reading

• Which Quadrant Is Your Product In? Product AI Depth vs. Commercialization Height
• In the Eye of the Storm, Return to Essentials | Hengshi’s Thinking and Action in the AI Era
• No Need to Pick Morning Blossoms at Dusk: The Present Is in Bloom | Hengshi February Headlines

About Hengshi

Hengshi is committed to defining the next generation of Agentic BI in the Data + AI era. Its product HENGSHI SENSE serves as an enterprise-grade data intelligence engine, helping customers and partners build vertical-domain AI analytics agent applications without code.

Founded in 2016, Hengshi is the only domestic data analytics platform vendor focused on enabling enterprise application SaaS partners. Hengshi serves large enterprise customers including WPP, BMW, GAC Honda, Publicis Groupe, BlueFocus, Sinopharm, AWS, Totole, and Genki Forest. It has also built deep partnerships with more than 200 enterprise application SaaS companies, including Phoenix Contact, Synagie in Singapore, TravelSky, Sangfor, Kingdee Cloud, Inspur Cloud, Seeyon, Ti-Net, Baozun E-commerce, FXiaoke, EC, Mingdao Cloud, and Beisen Kuxueyuan, enabling built-in business intelligence to go live immediately.