Scale Protocol: Feature rich platform designed to 10x you

Interact with an AI assistant that understands your codebase and can answer questions about it. The chat supports various commands for focusing on specific files, explaining code, and generating new code.

Get AI-powered code completions that understand the context of your code, going beyond simple IntelliSense suggestions to provide more meaningful and accurate completions.

Receive suggestions for what to edit next based on your current code and context, such as adding missing error handling, implementing TODOs, or refactoring opportunities.

Let the AI Agent perform complex coding tasks for you. The agent works through tasks step by step, showing its reasoning and asking for confirmation before making changes.

Get deep insights into your codebase with advanced code analysis, including complexity metrics, potential issues, dependency analysis, and symbol extraction.

Visualize the dependencies in your codebase to understand how different components relate to each other, with interactive graphs that you can zoom, pan, and filter.

Access a maintained knowledge base of your codebase, which Scale uses to provide more accurate assistance. You can view indexed files and symbols, add custom notes, and export the knowledge base.

Configure role-based access control to manage who can use different features, with the ability to create and manage roles, assign permissions, and view the permission matrix.

Access comprehensive audit logging for enterprise compliance, tracking API key usage, feature access, code sent to language models, and configuration changes.

View analytics to understand how your team is using the extension, including usage by feature and user, token consumption, and cost estimates.

Configure Single Sign-On for seamless authentication, with support for Azure AD, Okta, Google Workspace, and GitHub.

Fine-tune language models on your codebase for improved performance, with the ability to create fine-tuning datasets, start jobs, monitor progress, and deploy fine-tuned models.