Zoho announces major enhancements to Catalyst by Zoho, its platform as a service (PaaS), bringing new agentic development capabilities directly into developer programming environments.
New additions include Agent Skills, a non-interactive command line interface (CLI), and support for Model Context Protocol (MCP), plus new integrations with agentic AI-based programming assistants such as Anthropic’s Claude Code and OpenAI’s Codex.
Database and AI tools
Designed for both system integrators and software companies developing complex solutions, as well as independent developers who want to move from prototype to production without having to manage the infrastructure themselves, Catalyst by Zoho now also offers a free program for students that includes full-stack hosting, functions, database and AI tools, with the aim of facilitating access to these technologies and supporting new talent.
“In a very short time, AI programming assistants like Claude Code have transformed the speed at which applications are created, but getting that code into production and getting it to work correctly remains one of the main challenges. Developers still spend too much time connecting different cloud services and managing infrastructure. Catalyst by Zoho eliminates that complexity. By offering an agent-ready full-stack serverless platform, Catalyst enables AI developers and agents to create, deploy and scale intelligent applications independently of cloud providers and with much lower operational burden,” said Sridhar Iyengar, general manager of Zoho in Europe.
From AI-generated code to production applications
AI programming assistants can generate code quickly, but getting it to production requires in-depth knowledge of the platform, cloud services, and deployment processes. Catalyst brings all of these elements together into a single full-stack serverless platform, giving AI developers the capabilities they need to build, test, and deploy applications.
Catalyst Agent Skills provide programming assistants with the context necessary to understand Catalyst services, architecture, and recommended development patterns. Instead of letting the model decide on its own how to use Catalyst, the Skill guides the wizard toward the most appropriate tools and workflows. By displaying only the capabilities relevant to each task, it helps attendees choose the best tools and services, while optimizing token usage and getting accurate, verified results, even with lighter models.
The non-interactive CLI enables AI programming assistants to run multi-step workflows in Catalyst from start to finish without requiring human intervention at each stage. This reduces manual tasks and allows you to move faster from development to production.
Access to Catalyst capabilities
The Catalyst MCP Server allows AI programming assistants to directly access Catalyst capabilities from the developer’s regular work environment. Actions like creating a table in a database or adding a column can be done directly from VS Code, Claude Code, Cursor, or any AI-enabled IDE, without needing to switch to the Catalyst console. In this way, the entire development process is maintained within the same workflow and the launch of the application into production is accelerated.
Orchestration, integrated into the Skill, connects these three capabilities. When an AI programming assistant needs to decide how to execute a request, the Skill directs them to the CLI or MCP, rather than leaving that choice up to the model. This approach simplifies the developer’s work, reduces the risk of choosing an inappropriate tool, and helps generate more complete applications ready for production.
Zoho platform with human supervision
Although each service can function successfully independently, managing a fragmented technology infrastructure can increase operational complexity and hinder security, privacy, and governance. Applications built on Catalyst run in Zoho’s own data centers and are supported by the company’s security infrastructure, which includes DDoS protection, SOC compliance, VAPT, web application firewall, and other measures.
Human oversight is built into the deployment process itself, rather than added later, allowing organizations to maintain a greater level of control as AI becomes part of application development.
Separating development and production environments ensures that code is only pushed to production through manual approval. In this way, the AI agent does not directly access the production environment and the risk of errors resulting from autonomous intervention is reduced.
Access and permission controls determine who – or what – can participate in the development and what actions they can take.
Complete activity traceability makes it easy to monitor application logs, platform logs, and tool calls through MCP. From this data, developers can know precisely what actions the AI agent performed and when. All changes made after release are versioned, associated with their author, and can be reverted.
Human supervision is integrated into the deployment process itself, rather than added later
Catalyst is moving towards a new stage of agentic software development, in which developers will be able to collaborate with increasingly capable AI agents to execute complex workflows throughout the entire software development lifecycle, on a full-stack cloud infrastructure with integrated governance. Upcoming features include multi-agent support, connectors for AI tools, full-lifecycle agentic software development capabilities, and other features.
The Zoho platform maintains a simple, pay-as-you-go pricing model, where each feature costs per unit, without adding an additional license fee on consumption. Developers can check their usage level at any time from the Catalyst console.
