I Built LineLoom: A Local-First VS Code Autocomplete Extension — v1.0.0 Is Here

Write less. Build more. Stay in control.

Today, I'm excited to announce the first stable release of LineLoom v1.0.0, an open-source Visual Studio Code extension designed to provide inline code suggestions using locally configured Ollama models.

After progressing through multiple preview releases, LineLoom has reached an important milestone: it is now available through the Visual Studio Marketplace, Open VSX Registry, and GitHub Releases.

Why I started LineLoom

Coding assistants can be useful, but I wanted to explore an approach where developers have more control over the models they use and the context they share.

The goal behind LineLoom was to build a focused autocomplete extension rather than a full autonomous coding agent.

The design prioritizes native VS Code integration, configurable models, privacy-aware behavior, and an open-source implementation.

What LineLoom offers

1. Native inline code suggestions

LineLoom uses VS Code's inline completion system to offer suggestions directly in the editor.

It supports automatic and manual triggering, multiline suggestions, cancellation of outdated requests, and configurable generation settings.

2. Ollama integration

Developers can use a compatible Ollama model that they install and configure independently.

LineLoom also allows per-language model selection, making it possible to choose different installed models for different programming languages.

3. Nine language IDs

The initial stable release includes support for:

Python, JavaScript, TypeScript, Rust, and Kotlin.

It also registers experimental support for Go, Java, C, and C++.

4. Privacy-aware design

LineLoom operates locally by default and doesn't include built-in telemetry.

It uses bounded context from the active document rather than indexing an entire repository.

Trusted workspace checks, ignore rules, sensitive-file exclusions, and remote-endpoint approval controls help users manage which source context is sent to an inference endpoint.

When a user explicitly configures and approves an external HTTPS endpoint, the selected code context can leave the local machine. Local-first does not mean that all configurations remain offline.

5. More control over autocomplete

LineLoom includes Quick Actions for enabling and disabling suggestions, pausing automatic generation, selecting models, managing supported languages, checking the Ollama connection, and opening diagnostics.

It can also reuse the remainder of a recently generated suggestion when the user types a matching prefix under appropriate editor and privacy conditions.

How to install LineLoom

The easiest option is to open the Extensions panel in VS Code and search for LineLoom.

Verify that the publisher is SanskarCodes.

You can also install it using the VS Code CLI:

code --install-extension SanskarCodes.lineloom

Next, install and start Ollama independently and pull a compatible coding model. For example:

ollama pull qwen2.5-coder:1.5b

Open a trusted source-code workspace and use LineLoom: Quick Actions to select the model and check the connection.

Installation instructions are available on the official LineLoom website.

What does the stable release mean?

v1.0.0 is LineLoom's first stable distribution milestone.

The project includes automated TypeScript and lint checks, unit tests, packaging validation, and a VS Code extension-host integration test using a mocked Ollama server.

The GitHub release also includes a VSIX package and checksum.

However, these checks do not establish real-model completion quality, hardware performance, or full manual testing across Windows, Linux, and macOS.

Those areas remain open for additional testing and community feedback.

Where to get LineLoom

Visual Studio Marketplace:

https://marketplace.visualstudio.com/items?itemName=SanskarCodes.lineloom

Open VSX Registry:

https://open-vsx.org/extension/SanskarCodes/lineloom

GitHub repository:

https://github.com/sanskarIN/LineLoom

Stable release:

https://github.com/sanskarIN/LineLoom/releases/tag/v1.0.0

Official website:

https://sanskarin.github.io/LineLoom/

What's next?

I plan to continue improving completion behavior, documentation, testing, model compatibility, and developer experience.

Real-world bug reports and compatibility feedback are especially valuable as the project moves forward.

LineLoom is free and MIT-licensed. Developers are welcome to inspect the source, report issues, suggest improvements, and contribute.

If you're interested in local-first developer tools, I'd love to hear what you think.

Your model. Your editor. Your control.

Thanks for checking out LineLoom! 🚀