Table of contents
December brought a major leap forward for developers working with AI-generated code.
Tabnine launched Provenance and Attribution, a breakthrough feature that traces AI-generated code against GitHub’s public repositories, providing source repos and license information to help you make informed decisions about code implementation.
For development teams looking to standardize their AI workflows, we’ve introduced shared custom commands through a centralized .tabnine_commands file, enabling consistent command usage across your entire codebase.
We’ve also enhanced the developer experience with granular control over how you interact with AI Chat. Whether you prefer architectural discussions focused on system design patterns or step-by-step guidance for complex implementations, our new custom chat behaviors adapt to your working style. Plus, with configurable response lengths, you can toggle between concise answers for quick solutions and comprehensive explanations for deeper understanding.
We recently introduced Provenance and Attribution, a new capability that allows Tabnine users to get the benefits of the strongest LLMs out in the market while minimizing the risk of IP liability. Tabnine checks the code generated by its AI chat against all the publicly visible code on GitHub, flags any matches it finds, and provides the source repo as well as its license information. Tabnine also keeps a log of all the sources provided that Tabnine Admins can review. This capability makes it simpler for customers to review the suggestions and determine if they meet their requirements.
This feature is currently available as a private preview for enterprise customers. Reach out to our Support team to enable.
Users can specify Tabnine AI Chat to behave in certain ways (e.g., “Respond like a mentor with step-by-step instructions and examples” or “Respond in German/Spanish”). With custom chat behaviors, developers can mold Tabnine to match their preferred learning and working styles — whether that’s detailed step-by-step guidance for complex implementations or quick responses in their native language.
For example, Tabnine users can focus on “code style,” asking the Tabnine AI Chat to “Answer like an architect emphasizing scalability and system design patterns” or “Respond like a code reviewer focusing on best practices and potential optimizations.” Or if the developers are more junior, they can use a learning style-focused behavior, like “Answer like a CS professor with theoretical foundations and practical applications” or “Respond like a debugging coach who guides through problem-solving rather than giving direct solutions.”All of these custom behaviors will be considered as context when Tabnine delivers responses within AI chat.
Tabnine users can now customize their Tabnine AI Chat experience even further, allowing for more control over chat response length and style. Users can pick between “Concise” for a shorter answer and “Comprehensive” for a longer, further explanation of your prompts in chat.
This was a highly requested feature designed to enhance team collaboration and consistency. With this update, teams can now create, manage, and share custom commands across their projects, ensuring that all developers are working with the same set and version of commands. These capabilities ensure that every interaction with Tabnine is precisely tailored to both individual preferences and team-wide standards, making AI assistance more intuitive, efficient, and valuable for your entire development organization.
This feature is powered by a shared file, named .tabnine_commands, within each repository. Users can easily create, edit, and delete commands through the new interface in the chat settings. The shared file acts as the single source of truth for commands, making collaboration more seamless and efficient. Each team will be responsible for managing this file. At this time, shared commands do not support mentions.
Shared custom commands are now available for all supported IDEs for enterprise users.
Tabnine Enterprise self-hosted customers can now connect to their repositories via HTTPS access tokens. HTTPS credentials are generally simpler to set up and manage, and work better with credential managers built into operating systems.
Firewall compatibility is another significant advantage of HTTPS authentication. HTTPS traffic, which operates on port 443, is almost always allowed through corporate firewalls, while SSH (port 22) is often blocked in corporate environments for security reasons. This means teams can avoid the need to configure special firewall rules, simplifying their security infrastructure. Furthermore, HTTPS authentication often integrates more smoothly with other development tools and CI/CD pipelines, as many of these systems are designed with HTTPS in mind. This also extends to better compatibility with web-based interfaces and simpler integration with existing authentication systems.
Although SSH keys do offer certain advantages, such as better security through asymmetric encryption and eliminating the need to store passwords, the choice between HTTPS and SSH often depends on specific organizational needs and infrastructure requirements. That said, we’re happy to now offer both to our Tabnine Enterprise self-hosted customers.
These December updates reflect our focus on making AI a more pragmatic, secure, and customizable part of your development workflow. The combination of Provenance and Attribution for code traceability, shared custom commands for standardized team practices, and flexible chat interactions creates a more robust development environment that adapts to both individual coding styles and team-wide standards.
Enterprise self-hosted users will particularly benefit from our new HTTPS access token support, simplifying repository connections and improving compatibility with existing security infrastructure and CI/CD pipelines. As we continue to evolve Tabnine, we’re committed to building features that enhance your development workflow while maintaining the security and collaboration standards modern development teams require.
Ready to explore these new capabilities? Enterprise users can reach out to our Support team to enable the Provenance and Attribution preview, and all developers can start using custom chat behaviors and response lengths in their current Tabnine installation.
Want to see these new features in action and learn how to get more from Tabnine? Join our Tabnine Office Hours live demo and Q&A every Wednesday.