TabNin, provider of the code assistant of the same name, has announced a new AI agent specifically for code reviews. The purpose of this system is to help improve code quality, security, and compliance.
Advertisement
Natural language makes code standards accessible
Development teams can provide their standards in natural language to the system, for example, “Never send sensitive information (such as username or password) in URLs or query parameters, as these are not secure.” is recommended.”
Alternatively, users can also provide an existing repository from which TabNin’s AI retrieves specific rules by analyzing and annotating patterns. This means that built-in specifications can also be used for code review. The rules created by the system are also available in natural language and can be further customized, activated or deactivated for greater control by software developers.
Tabnine creates code review rules in natural language, making it easy to understand.
(Image: tabnine.com)
In addition to individually defined rules, Tabnine also offers predefined rules that can be activated when required. This includes general industry standards as well as best practices tailored specifically to programming languages and frameworks.
The system then automatically examines the code in the development environment and pull requests, detects deviations, and provides suggested solutions. Since the agent records codes contextually, i.e. not just as a sequence of characters or syntactic units, static code analysis should have fewer false alarms than traditional review systems.
Private beta for enterprise customers
TabNin’s new code review agent is available as an on-demand private beta to customers on the company’s Enterprise plan. Interested users can also use this Documentation for the new system Learn more, or sign up for one Livestream on November 7th Register where the company would like to present the technology.
(map)
CLC 2024: 5 workshops with expert knowledge on dev(s)operations and platform engineering
