LineLoom v1.2.0 — Stable Control Center and Local Completion

LineLoom v1.2.0 — GitHub Stable Release Notes

Channel: Stable GitHub release (not pre-release).
Publisher: SanskarCodes · Extension ID: SanskarCodes.lineloom ·
License: MIT · Minimum VS Code: 1.96

v1.2.0 is distributed as a stable-channel VSIX through GitHub Releases,
not as a pre-release build. The earlier
GitHub v1.1.1 pre-release
and v1.0.0 stable
are preserved. The GitHub release pipeline re-runs all automated checks,
verifies the VSIX is not marked pre-release, smoke-installs it and attaches
both the package and its SHA-256 checksum.

Manual QA not yet verified: Real-model Ollama completion quality and
CPU/GPU latency, physical Windows/macOS/Linux editor acceptance, hands-on
accessibility and privacy testing, and upgrades from earlier installed
versions have not been demonstrated. This release is owner-authorized for
GitHub distribution with those limitations disclosed, not represented as
passed. Those tests remain tracked in
issue #21.

Included capabilities

• Native LineLoom Activity Bar sidebar with completion, default/overridden
  models, enabled languages, active-editor context, privacy and connection status.
• Additional privacy-preserving in-memory completion engine status and counters:
  requests, suggestions, cancellations, cache hits, continuation reuse, empty
  results and failures. Average measured latency is shown only after the user
  enables performance diagnostics and a successful sample exists.
• Privacy-safe, categorical error messages for unavailable Ollama servers, missing models, timeouts, unauthorized requests and malformed/oversized responses, without exposing raw server error content.
• Model selection rejects duplicate, malformed, control-character-bearing and oversized model names returned from an untrusted Ollama endpoint; display is bounded to 512 entries. It also discards obsolete discovery results after endpoint/approval changes.
• Native inline suggestions via a configured Ollama model with manual/automatic
  triggering, multiline support, response filtering, cancellation and caching.
• Nine language IDs: Python, JavaScript (+JSX), TypeScript (+TSX), Rust, Kotlin,
  Go, Java, C, C++. The last four remain experimental. C# and Dart are
  not implemented and are not advertised as supported in v1.2.0.
• Workspace-trust restriction, path and ignore exclusions, blocked unapproved
  remote HTTPS inference, file-only document schemes and opt-in explicit remote
  approval. No built-in telemetry, remote account or cross-file indexing.

Packaging and installation

Install from this GitHub release:

1. Download LineLoom-1.2.0.vsix and LineLoom-1.2.0.sha256 from this release's assets.
2. Verify SHA-256 locally.
3. Open VS Code → Extensions → … → Install from VSIX…, not Visual Studio's
   separate VSIX Installer. Or use code --install-extension LineLoom-1.2.0.vsix.
4. Install Ollama and the selected coding model yourself. LineLoom never
   automatically installs models or starts sending source to remote servers.

Verification and known QA limitations

Automated evidence

The v1.2.0 pipeline checks cross-OS GitHub CI, typecheck, lint, unit tests,
static website verification, bundled extension build, mocked Ollama integration,
dependency audit, packaged VSIX identity and absence of VS Code pre-release
metadata, checksum generation, and isolated VS Code CLI installation.

Automated CI outcome: PASS. Windows/macOS/Linux build/test checks,
Linux mocked Ollama extension-host integration, and the stable-channel
VSIX archive/integrity/install smoke checks passed in
main-branch CI run #39366265411.
The dependency security audit also passed. The run generated and verified a
lineloom-v1.2.0-stable-candidate artifact. The separate GitHub release workflow
rebuilds and verifies the final stable VSIX before attaching public release assets.

No real-model performance measurements or hands-on acceptance can be inferred
from these checks.

Manual and real-model acceptance

Status: NOT TESTED / NO EVIDENCE PROVIDED for real Ollama quality and
CPU/GPU latency, genuine completion acceptance, interactive keyboard and
screen-reader/high-contrast sidebar testing, Windows/macOS/Linux physical editor
sessions, protected-file and remote approval tests with actual settings, and
upgrades from v1.0.0 stable and v1.1.1 preview.

Do not infer these passed from CI. Issue #21
tracks the mandatory release gates and
manual QA matrix records exact scenarios.

Distribution

This version is published to GitHub as a stable-channel release by explicit
owner request, with missing hands-on QA disclosed above. The
guarded GitHub release workflow
rechecks CI, the package, checksum, isolated installation and duplicate tag
protection. GitHub publication is separate from publishing to both external
registries.

• Visual Studio Marketplace
• Open VSX

Both listings must be independently updated and verified by the publisher.
An .exe is neither necessary nor appropriate for this cross-platform VS Code
extension: the installable executable extension artifact is a .vsix.

Support and bugs: https://github.com/sanskarIN/LineLoom/issues

LineLoom v1.1.1 — Active Editor Sidebar Preview

LineLoom v1.1.1 — Active Editor Sidebar Preview

GitHub pre-release · Publisher: SanskarCodes · MIT · VS Code 1.96+

This is an opt-in test build, following v1.1.0. The latest stable release
remains v1.0.0 on Visual Studio Marketplace,
Open VSX, and GitHub.
The planned v1.2.0 stable release is not yet published.

What's new since the v1.1.0 preview

• Active Editor group in the LineLoom Activity Bar Control Center displays
  the current supported programming language, effective configured Ollama model
  (global default or per-language override), and preliminary completion status.
• Explains when suggestions are unavailable due to missing/unsupported editors,
  disabled languages, disabled autocomplete, restricted workspaces, invalid
  endpoints, disabled remote endpoints, or unapproved remote HTTPS inference.
• Recognizes JavaScript React (JSX) and TypeScript React (TSX) language IDs.
• Matches the inline completion provider's file-only document scheme:
  untitled/virtual editors are not incorrectly presented as eligible.
• Refreshes sidebar status when switching active editors.
• Adds explicit accessibility labels to native Tree View sections and rows,
  with expanded unit, static privacy-contract and extension-host coverage.
• All sidebar state is derived without reading document contents, filenames,
  file paths or endpoint URL strings, and without making model requests.

How to try it

Download LineLoom-1.1.1.vsix from the assets below. Install in Visual
Studio Code → Extensions → … → Install from VSIX…. Windows users should not
double-click the VSIX in File Explorer: the Visual Studio VSIX Installer handles
a different extension format.

Or run code --install-extension LineLoom-1.1.1.vsix from the download directory.
Then choose the LineLoom icon in the Activity Bar or run
LineLoom: Open Control Center from the Command Palette.

Ollama and an appropriate coding model must be installed separately. Installing
the extension and opening the sidebar do not download or contact any model.

The companion LineLoom-1.1.1.sha256 asset verifies VSIX integrity. The package
uses VS Code's pre-release VSIX flag, not the stable-channel flag.

Important limitations

• Completion eligibility is settings-level only: an editor's displayed
  "Allowed by settings" does not bypass sensitive file, ignore-rule, or
  other request-time privacy gates. File permissions are enforced immediately
  before inference by the completion engine.
• Real Ollama model quality and CPU/GPU latency are not independently measured;
  manual VS Code UI, accessibility, workspace trust, and cross-platform user
  acceptance remain pending. No invented performance results are reported.
• Go, Java, C and C++ integration remains experimental. Cross-file repository
  indexing is not implemented. No built-in telemetry.
• Remote HTTPS endpoints require an explicit configuration opt-in and approval;
  loopback remains the default, and UI rendering does not contact the server.
• This GitHub pre-release is not automatically published to Microsoft
  Marketplace or Open VSX. Their stable v1.0.0 installation remains supported.

Next milestone: v1.2.0 stable

We will evaluate a stable package only after a documented manual and real-model
acceptance run, complete regression testing, and independent release checks.
See v1.2.0 plan
and release gate #21.

Repository: https://github.com/sanskarIN/LineLoom

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

Report problems: https://github.com/sanskarIN/LineLoom/issues

LineLoom v1.1.0 — Sidebar Preview

LineLoom v1.1.0 — Sidebar Preview

GitHub pre-release · Publisher: SanskarCodes · MIT · VS Code 1.96+

This is an opt-in test build of the next LineLoom release. The previously published
v1.0.0 remains the recommended stable channel on the Visual Studio Marketplace
and Open VSX. This release does not automatically publish to either registry.

New in this preview

• Native VS Code Activity Bar Control Center: discoverable LineLoom sidebar built
  with a native Tree View instead of a webview.
• Completion controls: enable/disable suggestions, pause automatic inference,
  trigger a manual suggestion and open the existing Quick Actions menu.
• Model & language controls: view configured model and override counts and invoke
  the existing approved model/language selectors.
• Privacy & connection display: see workspace trust and endpoint status by
  category. The sidebar never exposes the endpoint URL, editor source, prompts or
  file paths, and opening or refreshing it does not query Ollama.
• Session activity: aggregate request/suggestion/failure counters and a
  shortcut to the existing privacy-preserving diagnostics.
• Publisher settings fix: both Open Settings and the Setup Guide now target
  SanskarCodes.lineloom, not the old GitHub-owner-based identifier.
• Additional unit, integration, manifest, VSIX icon packaging and documentation
  checks for the sidebar.

Download and install

Download LineLoom-1.1.0.vsix from this GitHub release's assets, then:

1. Open Visual Studio Code → Extensions → … → Install from VSIX…
2. Select the downloaded .vsix; do not double-click it in Windows Explorer
   (that may invoke Visual Studio's incompatible VSIX Installer).
3. Restart/reload the extension host if prompted; use the new LineLoom Activity
   Bar icon, or LineLoom: Open Control Center in the Command Palette.
4. Independently install and start Ollama and an appropriate local coding model;
   no model download occurs when LineLoom is installed or the sidebar opens.

For CLI installation: code --install-extension LineLoom-1.1.0.vsix.
The companion LineLoom-1.1.0.sha256 is a checksum for download verification.

Security and known limitations

• Remote HTTPS endpoints remain off by default, requiring explicit opt-in
  and origin approval. Local loopback is the default.
• Inline completions remain restricted by workspace trust, ignore rules,
  sensitive-file exclusions and active-editor checks.
• No cross-file repository indexing and no built-in telemetry.
• Tests with mocked Ollama, typechecking, lint, website verification, CI host
  checks and package-install smoke tests are not a substitute for real model
  acceptance, manual sidebar accessibility checks or CPU/GPU latency benchmarks.
• Go, Java, C and C++ language support remains experimental.
• Existing v1.0.0 stable packages at Visual Studio Marketplace and Open VSX
  remain independent of this GitHub-only preview. Do not claim these registries
  have v1.1.0 until their public listings verify it.

Feedback: https://github.com/sanskarIN/LineLoom/issues

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

Follow the sidebar guide
and release acceptance checklist.

LineLoom v1.0.0 — Stable Release: VS Code + Ollama Inline Completions

LineLoom v1.0.0 — First Stable GitHub Release

Stable GitHub release · Publisher: SanskarCodes · MIT license · VS Code 1.96+

LineLoom is a VS Code extension that offers context-aware inline suggestions from your configured Ollama model. It is local-first by default, does not require a LineLoom account or a hosted inference subscription, and has no built-in telemetry. This stable channel is the first major-version distribution milestone following the v0.9.x public previews.

Core functionality

• Native VS Code inline suggestions with automatic and manual triggering, multiline output, configurable timeouts, and cancellation of stale requests.
• Ollama-compatible model selection with optional per-language overrides, model connection checks, editor status and Quick Actions.
• Language IDs: Python, JavaScript, TypeScript, Rust and Kotlin plus experimental Go, Java, C and C++.
• Privacy controls: trusted workspaces, sensitive/ignored-file exclusions, remote HTTPS endpoints disabled by default and explicit approval before sending context to a remote host.
• Bounded active-document context (no cross-file repository indexing), conservative output filtering, hashed suggestion caching, continuation, and optional aggregate timing diagnostics.
• Open-source development resources, GitHub Pages installation guide and test suites.

Changes for v1.0.0

• Stable package version and matching npm lockfile; publisher remains SanskarCodes, so the extension ID is SanskarCodes.lineloom.
• Release packaging is stable-channel VSIX, with no --pre-release flag. Preview packaging remains available separately using npm run package:preview.
• CI verifies TypeScript, lint, unit tests, static website, Linux VS Code extension-host integration with a mocked Ollama server, and packaged VSIX installation via a clean VS Code CLI.
• GitHub release workflow additionally requires npm audit --audit-level=high, ZIP/manifest identity verification, version match and an attached SHA-256 checksum.
• Website, installation instructions and changelog updated for the stable channel. Visual Studio Marketplace and Open VSX require separate authenticated publication and are not automatically populated by this release.

Download and install

1. Download LineLoom-1.0.0.vsix and (optionally) LineLoom-1.0.0.sha256 from this GitHub release.
2. In Visual Studio Code select Extensions → ... → Install from VSIX... and choose the VSIX. On Windows, do not double-click the file, because the full Visual Studio VSIX Installer is a different application.
3. From a terminal with VS Code's code command configured, an alternative is code --install-extension LineLoom-1.0.0.vsix in the download directory.
4. Install and start Ollama independently and obtain a compatible model, e.g. ollama pull qwen2.5-coder:1.5b.
5. In a trusted local source workspace, open LineLoom: Quick Actions, select your installed model, check the endpoint, then trigger a completion.

Website: https://sanskarin.github.io/LineLoom/
Installation help: https://sanskarin.github.io/LineLoom/install.html

What the checks establish — and what they do not

Automated tests validate a packaged VSIX that can be installed and listed by a real VS Code CLI, and test inline provider behavior against a mock Ollama HTTP server. These checks do not prove code accuracy or hands-on ghost-text acceptance using real models on every machine. One earlier user screenshot verified a v0.9.4 Windows install, not real completions with v1.0.0. No real CPU/GPU latency or complete offline QA measurements have been provided, so no such results are claimed. Report issues if you encounter editor or model problems.

Known limitations: Go, Java, C and C++ are experimental. No cross-file search or indexing. Sensitive-path rules are not full semantic secret detection. A user-approved external HTTPS endpoint receives bounded editor context and has its own privacy policy. Ollama/model availability, inference performance and output quality depend on local environment and chosen model.

The maintainer requested publishing the v1.0.0 stable tag despite the remaining non-blocking manual validation items. Follow the QA evidence matrix to continue real-device testing; do not present unchecked cases as verified.

Release scope: This is GitHub stable only. Visual Studio Marketplace and Open VSX publisher accounts, namespace approval and release uploads are separate follow-up work.

Source: https://github.com/sanskarIN/LineLoom
Issues: https://github.com/sanskarIN/LineLoom/issues
Publisher: Sanskar (SanskarCodes)
License: MIT for currently licensed source; historical separately distributed releases retain their original terms.

LineLoom v0.9.5 — VS Code Installation Smoke & Stable Readiness Preview

LineLoom v0.9.5 — VS Code Installation Smoke & Stable Readiness Preview

GitHub pre-release only · Publisher: SanskarCodes · MIT · Local-first code completion using Ollama

Why another preview?

The v0.9.4 user screenshot confirms LineLoom was installed and enabled in Visual Studio Code on Windows. The previous VSIX Installer dialog was Windows opening the downloaded VS Code package with the different full Visual Studio application. Installation alone does not confirm that real Ollama inference, accepted ghost text and privacy denial cases have worked.

This release adds stronger automated packaged-extension validation while keeping the final stable release blocked until actual user testing.

Changes since v0.9.4

• Add a real, isolated VS Code CLI installation smoke test in scripts/smoke-install-vsix.mjs. After archive validation, CI installs the VSIX using a clean downloaded VS Code and checks that SanskarCodes.lineloom appears at the exact version. The test profile is disposable and does not access personal workspaces or run models.
• Run the same packaged install smoke before GitHub publishes this pre-release. Existing Windows/macOS/Linux CI, Linux extension-host testing against mocked Ollama, privacy unit tests, security audit, archive identity validation, release checksums and pre-release flags remain in place.
• Update project website, installation page, README, changelog and publisher guides to v0.9.5.
• Add an explicit manual QA evidence matrix for real Rust/Python/JavaScript/TypeScript/Kotlin suggestions, editor accept behavior, lost Ollama connection, privacy boundaries, offline behavior and physical Windows/macOS/Linux testing.
• Preserve existing completion behavior; there is no new cloud service, mandatory account, model download, telemetry, or automatic Marketplace/Open VSX publication.

Install the preview

1. Download LineLoom-0.9.5.vsix from this release, plus LineLoom-0.9.5.sha256 to verify the download if desired.
2. In Visual Studio Code (not Visual Studio Community), open Extensions → … → Install from VSIX… and choose the file. On Windows, do not double-click .vsix in File Explorer.
3. Or from the folder containing the VSIX, run code --install-extension LineLoom-0.9.5.vsix if the code CLI is installed.
4. Start independently installed Ollama, install a compatible coding model if needed (for example ollama pull qwen2.5-coder:1.5b), then use LineLoom: Quick Actions to select it.
5. Test with a trusted scratch Rust/Python file. Use LineLoom: Trigger Completion, observe ghost text, and check actual acceptance.

Website: https://sanskarin.github.io/LineLoom/ · Installation guide: https://sanskarin.github.io/LineLoom/install.html

Known limitations and release gate

This is NOT stable v1.0.0. A successful headless CLI installation and a mocked HTTP response are not equivalent to manual editor keystrokes, model accuracy, practical inference latency, or complete cross-platform acceptance. At least one screenshot verifies an earlier v0.9.4 install but not real suggestions. Real CPU/GPU measurements and all required manual cases are not yet established.

Go, Java, C and C++ remain experimental. No cross-file indexing. Source context is restricted by trust and file exclusion rules. Remote HTTPS inference can send snippets to an explicitly approved external endpoint. Do not assume the extension detects all secrets in source content.

The stable v1.0.0 release and both external registry uploads (Visual Studio Marketplace and Open VSX) are reserved for a later, explicit user-approved final-release stage. GitHub's preview workflow does not upload to those registries.

Project: https://github.com/sanskarIN/LineLoom
Publisher ID: SanskarCodes · Display name: Sanskar
Issues: https://github.com/sanskarIN/LineLoom/issues

LineLoom v0.9.4 — VS Code Installation & VSIX Verification Preview

LineLoom v0.9.4 — VS Code Installation & VSIX Verification Preview

GitHub pre-release · MIT · Publisher: SanskarCodes · Requires VS Code 1.96+ and independently installed Ollama

Why this preview?

A user reported that double-clicking the v0.9.3 VSIX on Windows opened Visual Studio VSIX Installer and showed: “One or more extensions are for Visual Studio Code. Try installing them in Visual Studio Code.” That message comes from the wrong installer: the downloaded extension is for Visual Studio Code, not the full Visual Studio IDE. It does not by itself indicate a failed download, damaged VSIX, or incompatibility with VS Code.

Changes since v0.9.3

• Published a new step-by-step installation help page, covering Windows PowerShell and VS Code's built-in VSIX installation UI as well as macOS/Linux setup.
• Updated the homepage, README, publisher guide and troubleshooting documentation so users avoid file association mistakes.
• Added a read-only VSIX verification script that inspects ZIP structure and expected bundled extension files, confirms the packaged SanskarCodes.lineloom manifest identity/version matches the source, and rejects unexpectedly embedded dependencies, sensitive .env filenames and the standalone website.
• Both CI VSIX packaging and the GitHub preview release workflow now run the new archive verification before upload.
• Updated the manifest version/lockfile to 0.9.4 and pointed its homepage to the live GitHub Pages website.
• Preserved local-first Ollama completion, privacy controls, nine registered languages, MIT licensing, explicit --pre-release VSIX metadata and GitHub-only preview release policy.

Download and install: do not double-click on Windows

1. Download LineLoom-0.9.4.vsix from the GitHub pre-release assets.
2. Open Visual Studio Code (not Visual Studio Community/Professional/Enterprise).
3. Press Ctrl+Shift+X, click ... (Extensions menu), choose Install from VSIX…, and select the downloaded file.
4. Or in PowerShell, from the folder containing the file, run code --install-extension .\\LineLoom-0.9.4.vsix if your VS Code code CLI is configured.
5. Install/start Ollama and a coding model separately (example: ollama pull qwen2.5-coder:1.5b), open a trusted workspace, and run LineLoom: Quick Actions.

A genuine VS Code installation error should be reported with the exact VS Code error message (but no passwords, secrets, code snippets or personal paths). See the troubleshooting guide.

Verification and limitations

The release workflow requires a locked npm install, TypeScript, lint, unit tests, website checks, VS Code extension-host integration with a mocked Ollama endpoint, dependency audit, a pre-release-marked VSIX, read-only archive/manifest inspection, and SHA-256 checksum generation. GitHub CI checks Windows/macOS/Linux.

The verification script ensures the VSIX has the expected shape and publisher identity. It does not prove real installation or code-completion quality on each device. The stable v1.0.0 gate still requires manual VS Code installation tests, real Ollama CPU/GPU performance and model quality checks, and physical Windows/Linux/macOS testing. Go, Java, C and C++ remain experimental.

This is a GitHub pre-release, not a stable v1.0.0 release. Microsoft Visual Studio Marketplace and Open VSX require separate publisher-authorized uploads.

Website: https://sanskarin.github.io/LineLoom/
Documentation: https://sanskarin.github.io/LineLoom/install.html
Repository: https://github.com/sanskarIN/LineLoom

LineLoom v0.9.3 — GitHub Pages Website & Preview Distribution

LineLoom v0.9.3 — GitHub Pages Website & Preview Distribution

GitHub pre-release · MIT · publisher: SanskarCodes · local-first Ollama inline autocomplete

Changes since v0.9.2

• Added a dedicated, responsive static LineLoom project website under site/. Includes installation steps, transparent feature descriptions, supported languages, privacy and network behavior, common questions, GitHub release links, accessible navigation, and original SVG icons/social-preview artwork.
• Added the GitHub Pages deployment workflow using actions/configure-pages, actions/upload-pages-artifact and actions/deploy-pages. The site is expected at https://sanskarin.github.io/LineLoom/ once GitHub Pages is enabled with Source: GitHub Actions, and the deploy job and live URL are verified. Source code alone is not proof of a live website.
• Added Node.js static-site regression tests: release/version alignment, local asset existence, fragment links, responsive and reduced-motion styles, honest privacy/preview claims, and no invented registry listings. These run as part of npm run check, together with existing TypeScript/ESLint/unit/build verification.
• Excluded standalone site files from the VSIX with .vscodeignore, preserving a focused extension artifact.
• Preserved SanskarCodes.lineloom publisher identity, explicit VS Code --pre-release packaging, nine registered languages, local-first defaults and privacy safety gates; no new model provider or account requirement.

Download and install

1. Download LineLoom-0.9.3.vsix and optionally LineLoom-0.9.3.sha256 from this GitHub pre-release.
2. Use VS Code ^1.96.0. Choose Extensions → ··· → Install from VSIX, or code --install-extension LineLoom-0.9.3.vsix.
3. Install and start Ollama independently; download a compatible model manually (example ollama pull qwen2.5-coder:1.5b).
4. Open a trusted supported-language workspace. In the Command Palette run LineLoom: Quick Actions and verify the connection/model, then request code suggestions.
5. Check the VSIX checksum with sha256sum -c LineLoom-0.9.3.sha256 using a compatible shell.

Verified automation and outstanding QA

The pre-release workflow requires npm ci, TypeScript, lint, unit tests, static-site tests, a bundle build, Linux VS Code extension-host integration using mock Ollama, dependency audit, vsce --pre-release, VSIX ZIP integrity verification and SHA-256 generation. CI also checks Windows and macOS.

The website is a GitHub Pages deployment, not a hosted inference service. Initial GitHub Pages enablement may need an authorized repository owner to choose Settings → Pages → GitHub Actions. Consult deployment instructions.

Still unverified for stable v1.0.0: real model accuracy, offline onboarding on every platform, CPU/GPU inference benchmarks, manual ghost-text acceptance, per-language output quality and physical Windows/macOS/Linux smoke checks. Go, Java, C and C++ remain experimental.

GitHub pre-release does not automatically publish the VSIX to Visual Studio Marketplace or Open VSX. Publisher verification and external publication are separate operations.

License: current MIT project code; historical Apache-2.0 release copies retain their original terms. Independently installed models and dependencies keep their own licenses.

Repository: https://github.com/sanskarIN/LineLoom
Website (after successful Pages deployment): https://sanskarin.github.io/LineLoom/
Publisher: Sanskar (SanskarCodes)
Support: https://github.com/sanskarIN/LineLoom/issues

LineLoom v0.9.2 — Marketplace Identity & Preview Packaging

LineLoom v0.9.2 — Marketplace Identity & Preview Packaging

GitHub pre-release only · Publisher ID: SanskarCodes · MIT license · local-first VS Code code completion

What changed since v0.9.1

• Publisher identity: The extension publisher in package.json is now SanskarCodes, matching the user's Microsoft Visual Studio Marketplace publisher account. The extension identifier to use across registries is SanskarCodes.lineloom, while the GitHub repository remains sanskarIN/LineLoom.
• Pre-release VSIX: the default npm run package script and GitHub pre-release workflow now use the VS Code packaging tool's --pre-release flag. A GitHub pre-release label alone does not mark a VSIX as a Marketplace pre-release.
• Reproducibility: synchronized package and lockfile at 0.9.2, plus a test guarding publisher identity, version consistency and pre-release packaging commands.
• Automated verification: CI on Windows/macOS/Linux, Linux VS Code extension-host tests using a mocked Ollama server, VSIX integrity and SHA-256 checksums, and a pre-release-only GitHub workflow with explicit duplicate-tag protection.
• Publishing documentation: separate manual Marketplace/Open VSX onboarding instructions, accurate claims policy, and outstanding stable-release QA gates.
• No changes to inference semantics: existing nine registered languages, Ollama configuration, model switching, privacy gates, caching, Quick Actions, typed continuation, and remote endpoint consent are retained.

Download and install

1. Download LineLoom-0.9.2.vsix from this GitHub pre-release when the asset is published, plus the accompanying checksum file if desired.
2. Open compatible VS Code (^1.96.0) and choose Extensions -> ... -> Install from VSIX, or run code --install-extension LineLoom-0.9.2.vsix.
3. Install Ollama and a coding model independently, for example ollama pull qwen2.5-coder:1.5b; start the Ollama service.
4. Open a trusted source workspace, use LineLoom: Quick Actions to check the endpoint and model, and test suggestions with LineLoom: Trigger Completion.
5. Verify the file hash with sha256sum -c LineLoom-0.9.2.sha256 on a compatible shell.

Scope and limitations

This GitHub pre-release contains a pre-release-marked VSIX intended for user-controlled testing and subsequent registry uploads. It does not itself publish LineLoom to the Microsoft Marketplace or Open VSX. Both registries require authorized publisher operations and validation. See publishing guide.

The automated VS Code extension-host suite uses synthetic fixtures and a mocked inference server. Actual model quality, CPU/GPU inference latency, fully offline first-run setup, manual ghost-text acceptance, and hands-on testing of every OS/language remain unverified. Go, Java, C and C++ are experimental. Remote HTTPS inference can send allowed editor snippets to an approved external endpoint. There is no telemetry or cross-file source indexing.

This is not stable v1.0.0. The stable gate remains blocked.

License: MIT for material authorized for MIT distribution; historical Apache-2.0 releases retain their original license terms. See licensing.

Project: https://github.com/sanskarIN/LineLoom
Publisher brand: Sanskar (SanskarCodes)
Maintainer: https://github.com/sanskarIN
Optional support: https://buymeacoffee.com/SanskarCodes · https://razorpay.me/@SanskarCodes · https://paypal.me/SanskarCodes

LineLoom v0.9.1 — Pre-Dispatch Privacy & Model-Picker Safety Preview

LineLoom v0.9.1 — Pre-Dispatch Privacy & Model-Picker Safety Preview

GitHub pre-release · MIT license · local-first VS Code inline autocomplete using independently installed Ollama

Changes since v0.9.0

• Prevent stale requests before sending code: the extension rechecks the current editor, trusted workspace, active model/language settings, custom file exclusions and remote-server approval after the typing delay and private-file scans, immediately before contacting Ollama.
• Stricter cached suggestions: typed continuations and exact-result cache returns are also gated by up-to-date settings, rather than trusting only the configuration captured before debounce.
• Safer model pickers: LineLoom: Select Model and Assign Language Model discard choices retrieved from an endpoint that is no longer configured or approved. Approval dialogs no longer grant consent to an endpoint after the user switches it while the dialog is open.
• Additional configuration guard: changing excluded-path patterns or request timeout during a pending completion invalidates the old request.
• Regression tests: Windows, macOS and Linux CI checks, pure settings/approval tests and actual VS Code extension-host testing with a simulated Ollama server. In the editor tests, changing model or ignore patterns while an automatic completion waits to dispatch produces no Ollama generation request.
• Existing features retained: per-language model assignment, global fallback, nine registered languages (Go, Java, C and C++ experimental), typed continuation reuse, Quick Actions, local-first Ollama inference, conservative file exclusions, no telemetry, explicit HTTPS remote approval and HTTP redirect rejection.

No new runtime dependencies, source indexing, background model downloads, hosted service or account requirement have been introduced.

Install / upgrade

1. Use VS Code compatible with extension engine ^1.96.0, and install/start Ollama plus your chosen coding model yourself.
2. Download LineLoom-0.9.1.vsix from this GitHub pre-release's assets.
3. In VS Code choose Extensions → … → Install from VSIX, or run code --install-extension LineLoom-0.9.1.vsix.
4. Open a trusted source workspace and use LineLoom: Quick Actions to check the current server, select the global model and optionally assign per-language models.
5. Use LineLoom: Trigger Completion to request a suggestion. If settings or consent change during an inference request, request a fresh completion.
6. Optionally verify the LineLoom-0.9.1.sha256 asset with sha256sum -c LineLoom-0.9.1.sha256 on a supported shell.

Automated checks and limitations

The publication workflow requires a locked npm install, TypeScript typechecking, ESLint, Vitest, extension bundle, VS Code extension-host integration with mocked Ollama, npm dependency audit, VSIX ZIP integrity verification and SHA-256 generation. Separate repository CI runs on Windows, macOS and Ubuntu.

Still not verified: real Ollama model quality or latency, physical CPU/GPU performance, fully offline first-run setup, manual install and ghost-text acceptance on all three operating systems. Pre-dispatch revalidation prevents outdated requests from starting; it does not recall code previously sent to an approved endpoint. LineLoom does not scan file contents for secrets and does not perform cross-file context indexing.

This is a GitHub pre-release, not a stable v1.0.0 release, and there is no claimed VS Code Marketplace/Open VSX listing.

License: MIT for authorized project-owned material; historical Apache-2.0 releases keep their original terms. Dependencies and separately installed models are governed by their own licenses. https://github.com/sanskarIN/LineLoom/blob/main/docs/licensing.md

Repository: https://github.com/sanskarIN/LineLoom · Issues: https://github.com/sanskarIN/LineLoom/issues · Security: https://github.com/sanskarIN/LineLoom/blob/main/SECURITY.md

Maintainer: https://github.com/sanskarIN · https://x.com/SanskarCodes · https://linkedin.com/in/SanskarCodes

Optional support: https://buymeacoffee.com/SanskarCodes · https://razorpay.me/@SanskarCodes · https://paypal.me/SanskarCodes

LineLoom v0.9.0 — Per-Language Ollama Models Preview

LineLoom v0.9.0 — Per-Language Ollama Models Preview

GitHub pre-release · MIT-licensed LineLoom source · local-first VS Code inline autocomplete

Changes since v0.8.1

• Different models for different programming languages: set a custom installed Ollama model for Python, JavaScript/JSX, TypeScript/TSX, Rust, Kotlin, Go, Java, C or C++. Each language without an override continues to use the global default model.
• Quick Actions workflow: choose LineLoom: Assign Language Model from the status bar or Command Palette. Pick a language, then an installed model, or choose Use Default Model to clear its override.
• No automatic model downloads: the picker reads models from the already configured Ollama origin only after user initiation. A remote origin continues to require separately approved HTTPS access.
• Correct inference routing: model selection affects prompts (including supported FIM routing), Ollama request payloads, exact-result cache hashes and typed-continuation profile hashes. Changes to language-model preferences invalidate stale in-flight suggestions.
• Safety and tests: unknown keys, invalid model values and control characters in manual settings are ignored. The unit suite covers mapping and fallback; a VS Code extension-host integration test confirms the model field sent to a simulated Ollama server changes for TypeScript and returns to the global default when an override is cleared.
• Existing features remain: nine supported language IDs, Quick Actions, manual-only mode, conservative exclusion of private files, bounded active-document context, HTTP redirect rejection, per-origin remote HTTPS approval and no telemetry.

Installation and usage

1. Install VS Code compatible with extension engine ^1.96.0. Start Ollama and install the coding models you want independently.
2. Download LineLoom-0.9.0.vsix from this GitHub release's assets.
3. Install through Extensions → … → Install from VSIX, or run code --install-extension LineLoom-0.9.0.vsix.
4. Open a trusted workspace, click the LineLoom status-bar item and check your Ollama connection.
5. Choose Select Model for the global default, then Assign Language Model to configure optional per-language overrides. Model names can also be edited under the lineloom.languageModels setting.
6. Verify the attached SHA-256 checksum with sha256sum -c LineLoom-0.9.0.sha256 on supported systems.

Verification and limitations

Release verification includes locked npm ci, TypeScript, ESLint, Vitest unit tests, extension bundling, a real VS Code extension-host test against mocked Ollama, dependency audit, VSIX ZIP inspection and SHA-256 checksum generation. Repository CI separately checks Windows, macOS and Ubuntu.

Not yet verified: the usefulness or latency of installed real models across languages, comparative local CPU/GPU performance, completely offline end-user setup, and manual VSIX acceptance on each desktop operating system. The language-specific model picker does not rank models or guarantee they are optimized for a language. Go, Java, C and C++ remain experimental.

This is a GitHub pre-release, not stable v1.0.0. No VS Code Marketplace or Open VSX listing is claimed.

License: MIT for authorized project-owned LineLoom material; previous Apache-2.0 release tags retain their own license terms, and dependencies/models are separately licensed. https://github.com/sanskarIN/LineLoom/blob/main/docs/licensing.md

Source: https://github.com/sanskarIN/LineLoom · Issues: https://github.com/sanskarIN/LineLoom/issues · Security: https://github.com/sanskarIN/LineLoom/blob/main/SECURITY.md

Maintainer: https://github.com/sanskarIN · https://x.com/SanskarCodes · https://linkedin.com/in/SanskarCodes

Optional support: https://buymeacoffee.com/SanskarCodes · https://razorpay.me/@SanskarCodes · https://paypal.me/SanskarCodes