Trelent - AI Docstrings on Demand

Generates language-specific docstrings by analyzing the function signature and body at the current cursor position, then inserts the formatted docstring directly into the source file. The extension reads the active editor buffer, extracts the function context, sends it to a cloud-based AI backend, and receives a formatted docstring that matches the target language's standard (JSDoc for JavaScript, JavaDoc for Java, XML for C#, ReST/Google/Numpy for Python). Activation occurs via keyboard shortcut (Alt+D / Cmd+D) or context menu, making it an on-demand, synchronous operation integrated into the code editing workflow.

Automatically detects the file type of the active editor and generates docstrings conforming to that language's standard documentation format. For Python, the extension supports multiple formats (ReST, Google, Numpy) with format selection mechanism undisclosed; for JavaScript, Java, and C#, it generates JSDoc, JavaDoc, and XML formats respectively. The AI backend receives language context from the file extension and produces output matching the appropriate docstring syntax, including parameter descriptions, return type documentation, and exception handling where applicable.

Processes function code through a cloud-based AI backend (model architecture and provider undisclosed) that analyzes function signatures, parameter names, return types, and implementation logic to generate semantically appropriate docstrings. The backend stores anonymized source code for service improvement, meaning identifying information is stripped but code structure and logic patterns are retained. Communication is one-way: the extension sends code to the backend and receives generated docstring text; no iterative refinement or feedback loop is documented.

Integrates into VS Code's command palette, keyboard binding system, and right-click context menu to provide multiple activation paths for docstring generation. The primary shortcut is Alt+D (Windows/Linux) or Cmd+D (macOS), registered via VS Code's keybinding API. The extension also appears in the context menu when right-clicking in a text editor, allowing mouse-based activation. Activation is synchronous and cursor-position-aware: the extension reads the current cursor location, identifies the enclosing function, and triggers generation without requiring explicit function selection.

Analyzes the code at the current cursor position to identify the enclosing function, extract its signature (parameters, return type), and read its implementation body. The extension uses language-specific parsing (mechanism undisclosed) to determine function boundaries, parameter names, types, and return type information. This context is then sent to the AI backend for docstring generation. The extraction is scoped to the current function only; no cross-function or class-level analysis is performed.

Offers a free tier providing cloud-based docstring generation with anonymized data retention for model improvement, and an enterprise tier enabling self-hosted deployment on customer infrastructure. The free tier uses Trelent's cloud backend (no usage limits documented); the enterprise tier allows on-premises deployment with no data transmission to Trelent servers. Pricing details for enterprise are not published; interested customers must contact Trelent directly. The freemium model is designed to reduce friction for individual developers while offering privacy-preserving options for enterprises.

Explicitly disclaims 100% accuracy of generated docstrings and requires users to manually review all output before committing to version control or production. The extension does not provide built-in validation, linting, or comparison against the actual code; users must visually inspect generated docstrings for semantic correctness, parameter accuracy, and consistency with implementation. This design places responsibility on the user to catch errors, hallucinations, or misinterpretations by the AI backend.