GetBotAI Code assistant

Provides real-time code completion suggestions directly in the VS Code editor by routing user input to configurable AI models (GPT-4o, Claude Sonnet, DeepSeek, Gemini) via GetBotAI's backend API. The extension monitors cursor position and code context, sending the current file buffer and selection state to the inference backend, which returns completion suggestions rendered as inline autocomplete proposals. Supports model switching without extension reload, allowing developers to compare completion quality across providers.

Analyzes the current file or selected code block to identify syntax errors, logic bugs, and runtime issues by sending code to the configured AI model with error-detection prompts. The extension parses the AI response to extract identified issues and suggested fixes, presenting them in a structured format within the sidebar or chat interface. Developers can apply fixes with a single click, which replaces the problematic code block with the corrected version.

Implements usage-based rate limiting through GetBotAI's backend, with different query limits based on subscription tier (free trial: 3 days, Silver tier, Gold tier). Each API call to the backend consumes a query quota, and the extension tracks remaining quota in the UI. When quota is exhausted, the extension prevents further requests and prompts the user to upgrade or wait for quota reset.

Enables developers to create a single GetBotAI account that works across VS Code extension, Chrome browser extension, and Edge browser extension. Account credentials and custom commands/prompts are synchronized across platforms, allowing seamless switching between tools. The extension authenticates via email signup on the GetBotAI website and maintains session state across platforms.

Generates natural-language explanations of code functionality by sending the selected code block to the configured AI model with a structured explanation prompt. The model returns a description of what the code does, how it works, and why it's structured that way. Explanations are rendered in the chat sidebar with full conversation history, allowing developers to ask follow-up questions about specific parts of the explanation.

Analyzes selected code to identify optimization opportunities (performance bottlenecks, readability improvements, memory efficiency) by sending the code to the AI model with optimization-focused prompts. The model returns a prioritized list of suggested optimizations with explanations of performance impact and refactoring steps. Developers can review suggestions in the chat interface and apply recommended changes via inline code replacement.

Scans selected code for security vulnerabilities, specifically SQL injection risks and resource leak patterns, by sending code to the AI model with security-focused analysis prompts. The model identifies vulnerable code patterns (e.g., string concatenation in SQL queries, unclosed file handles) and suggests secure alternatives (parameterized queries, try-finally blocks). Results are presented as a prioritized vulnerability list with severity levels and remediation steps.

Analyzes code containing threading, async/await, or lock-based concurrency patterns to identify potential deadlock scenarios by sending code to the AI model with deadlock-detection prompts. The model identifies problematic patterns (circular lock dependencies, nested locks, missing timeouts) and suggests refactoring approaches (lock ordering, timeout mechanisms, lock-free data structures). Results include visual representations of lock dependency graphs and step-by-step deadlock scenarios.

Generates Cucumber step definitions and test scenarios from selected code by analyzing the function signature, parameters, and logic, then creating BDD-style test steps in Gherkin syntax. The extension sends code to the AI model with test-generation prompts, receives step definitions (Given/When/Then format) and corresponding implementation code, and presents them in the chat interface for review and insertion into test files. Supports multiple testing frameworks and languages.

Provides a conversational interface within the VS Code sidebar where developers can ask questions about code, request modifications, and receive explanations through multi-turn dialogue. The extension maintains conversation history within the session, allowing developers to reference previous questions and build on prior context without re-selecting code. Each message is routed to the configured AI model, and responses are rendered with syntax highlighting for code blocks and markdown formatting for explanations.

Allows developers to select and switch between 9+ AI models (GPT-4o, Claude Sonnet, DeepSeek, Gemini Flash, etc.) from different providers (OpenAI, Anthropic, Google, DeepSeek) without restarting the extension or losing context. Model selection is exposed in the extension settings and UI, with featured models (GPT-4o, Claude Sonnet, DeepSeek Reasoner) highlighted as recommended. Switching models applies to all subsequent requests within the session, enabling A/B comparison of model outputs.

Allows developers to create, save, and reuse custom commands and prompts within the extension, enabling workflow automation and standardization. Custom commands can be triggered from the command palette or context menu, executing predefined prompts with the selected code. The extension stores custom commands in a library accessible across sessions, supporting parameterization and templating for dynamic prompt generation.