Patch Release v0.8.1 📎

[tnaum-ms]

Patch Release v0.8.1

This patch delivers Node 24 compatibility, significantly faster extension startup, tree view quality-of-life improvements with live item counts, AI transparency in Query Insights, batch connection deletion, and a fantastic wave of community contributions.

🤲 Community Contributions

This release would not be what it is without our fantastic open-source community. A big thank you to everyone who submitted pull requests, reported issues, and helped shape v0.8.1!

• @lte-z contributed two improvements in this release: hidden index visibility and context menu cleanup (#674) and shard key display in collection tooltips (#670).
• @omribz156 added contextual filenames for new Query Playground files (#664).
• @Jacquelinezhong fixed the _id_ index sort order so it always appears first in the Indexes list (#662).
• @Green00101 cleaned up the obsolete release-notes notification migration code (#622).
• @CalvinMagezi, @Jah-yee, and @Enocko each independently submitted a fix for the credentialId → clusterId parameter rename (#568, #575, #651)! Three contributors tackling the same issue is a rare and wonderful thing. All of their work influenced the final merged result (#652).

What's Changed in v0.8.1

💠 Node 24 Compatibility (#699)

VS Code 1.123 (released June 3, 2026) ships with Node 24. If you updated VS Code and found the DocumentDB extension no longer loading, this fix is for you. The extension now works correctly with VS Code 1.123 and later.

💠 Item Counts on Tree Nodes (#714)

The tree view now tells you more before you click. Two new at-a-glance counts appear automatically as you browse:

• Collection count on database nodes: After expanding a cluster, each database shows how many collections it contains (e.g., ·· 12). For large clusters, counts cap at 50+ to keep the query lightweight.
• Index count on the Indexes folder: Expanding a collection now shows how many indexes are present at a glance (e.g., ·· 4).

Counts load asynchronously in the background so tree expansion remains instant. A new documentDB.accessibility.hideCountPrefix setting lets you suppress the ·· visual separator for a cleaner look or better screen reader experience.

💠 AI Model Transparency in Query Insights (#690)

Query Insights now tells you exactly which AI model analyzed your query and confirms upfront that the feature uses a utility model that does not count against your GitHub Copilot premium request quota. Both the pre-invocation card and the post-response panel now show:

• A persistent "No additional cost for most GitHub Copilot subscribers" disclosure with a Learn more link.
• A "Powered by {model} via GitHub Copilot" attribution byline after each successful analysis.

What is a utility model? Utility models are included in all paid GitHub Copilot plans at no extra charge (GPT-4o, GPT-4.1, and GPT-5 mini all carry a 0x request multiplier, meaning they do not consume your monthly premium request allowance). For the full breakdown by plan, see AI Features: Utility Model and Pricing in the user manual.

💠 Batch Connection Deletion (#667)

You can now select multiple connections in the Connections View and delete them all at once. The command adapts its confirmation message to reflect the number of selected items, continues through individual failures so a single bad connection does not block the rest, and reports a summary when done. Clearing out stale connections from a long list just became much less tedious.

💠 Faster Extension Startup and Connection Loading (#726)

The extension loads noticeably faster, with the biggest gains for users on Remote-WSL or anyone with a large number of saved connections. Your connections appear in the tree sooner after VS Code opens.

Note: This release removes the one-time import of connections from the legacy Azure Databases VS Code extension (ms-azuretools.vscode-cosmosdb). If you have connections in that extension that were never opened in DocumentDB for VS Code, you will need to re-add them manually using a connection string.

💠 Performance: Throttled Background Document-Count Fetches (#685)

When a database with many collections is expanded, every collection fires a background document-count request. Previously, these all launched in parallel, potentially saturating the connection pool and competing with foreground queries. A new per-cluster concurrency limiter caps simultaneous count fetches at 5 and staggers them 250 ms apart, dramatically reducing server load while keeping the tree feeling responsive.

💠 Hidden Index Visibility and Context Menu (#656, #674)

Hidden indexes are now clearly labeled in the tree view with a hidden description, and the right-click context menu has been tightened so only valid actions appear: hidden indexes show only Unhide Index…, non-hidden indexes show only Hide Index…, and the _id_ index shows neither. No more guessing which action applies.

⭐ Thanks to @lte-z for this contribution!

💠 Shard Key in Collection Tooltip (#661, #670)

Sharded collections now expose their shard key in the hover tooltip alongside document count and storage size. The information is extracted directly from the listCollections() response with no extra round-trips, and is only shown when a shard key is actually present; unsharded collections are unaffected.

⭐ Thanks to @lte-z for this contribution!

💠 Contextual Query Playground Filenames (#660, #664)

New Query Playground files opened from the tree now get meaningful names derived from their context — for example, myCluster_myCollection.documentdb.js instead of a generic untitled name. Invalid filename characters are stripped automatically, and a numeric suffix is appended if the same name is already open.

⭐ Thanks to @omribz156 for this contribution!

💠 _id_ Index Always Sorted First (#657, #662)

The _id_ index — present on every collection — now always appears at the top of the Indexes list regardless of other index names. Previously, indexes with uppercase-leading names could sort above _id_ due to locale comparison rules.

⭐ Thanks to @Jacquelinezhong for this contribution!

💠 Community Code Quality: credentialId → clusterId Rename (#567, #652)

The internal credentialId parameter in ClustersClient and ClusterSession has been renamed to clusterId to match the established naming convention throughout the codebase. This was a popular community contribution — three contributors submitted independent PRs for the same issue, and all of their work informed the final result.

⭐ Thanks to @CalvinMagezi (#568), @Jah-yee (#575), and @Enocko (#651) for independently contributing to this fix!

💠 Removed Obsolete Notification Migration Code (#611, #622)

Pre-0.7.0 transitional release-notes notification logic and a leftover 0.8.0-bugbash migration path have been cleaned out, simplifying the notification system.

⭐ Thanks to @Green00101 for this contribution!

💠 Dependency Updates (#654, #672, #673, #678)

Updated webpack-dev-server (5.2.3 → 5.2.4, includes a CORP security header fix), @nevware21/ts-utils (0.13.0 → 0.14.0), and qs/express to their latest versions.

Changelog

See the full changelog entry for this release:
➡️ CHANGELOG.md#081