BE-930: Add memory, thread and disk limits to the workspace

[thehabbos007]

🌟 Purpose

This PR lets a caller limit a workspace's memory, threads and disk use through WorkspaceConfig.

🔍 Changes

• memory_limit, temp_directory_limit and threads set DuckDB's memory_limit, max_temp_directory_size and threads settings.
• disk_limit limits the combined size of the database file, its log and DuckDB's temporary directory. The workspace checks it on open and after each query and create_table. When the files are over the limit, it checkpoints the log into the database file and measures again before it fails with DiskLimit.

⚠️ Known issues

• The check runs after a statement, so one create_table can write past the limit, and the table it created stays in the database.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
hash	Ready	Preview	Oct 13, 2026 9:22pm UTC
petrinaut	Ready	Preview	Oct 13, 2026 9:22pm UTC
petrinaut-docs	Ready	Preview	Oct 13, 2026 9:22pm UTC

1 Skipped Deployment

Project	Deployment	Actions	Updated
hashdotdesign-tokens	Ignored	Preview	Oct 13, 2026 9:22pm UTC

[codecov]

Codecov Report

❌ Patch coverage is 94.68085% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 66.78%. Comparing base (4341978) to head (b241522).

Files with missing lines	Patch %	Lines
...al/integrations/engine/src/workspace/connection.rs	93.97%	4 Missing and 1 partial ⚠️

Additional details and impacted files

@@                              Coverage Diff                               @@
##           asa/hash-integrations/04a-engine-snapshots   #10003      +/-   ##
==============================================================================
+ Coverage                                       66.77%   66.78%   +0.01%     
==============================================================================
  Files                                            1982     1983       +1     
  Lines                                          217586   217659      +73     
  Branches                                         8513     8520       +7     
==============================================================================
+ Hits                                           145297   145368      +71     
- Misses                                          70662    70663       +1     
- Partials                                         1627     1628       +1     

Flag	Coverage Δ
hash-integrations-engine	89.95% <94.68%> (+0.96%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[cursor]

PR Summary

Medium Risk
Changes workspace open API and enforces disk limits after queries; misconfigured limits or checkpoint failures could block opens or statements in production integrations.

Overview
Introduces WorkspaceConfig and a ByteSize helper so callers can cap DuckDB memory, temp directory size, thread count, and overall workspace disk use when opening a workspace. Workspace::open now takes (path, config) and applies the DuckDB settings on connect.

When disk_limit is set, the engine totals the database file, WAL, and .tmp directory (with recursive sizing), runs CHECKPOINT if over limit, then rejects open or post-query / create_table with DiskLimit / DiskUsage errors. A single statement can still grow files past the cap before the check; data written in that request is not rolled back.

Integration tests cover configured DuckDB settings, open refusal, post-statement limits, temp-dir accounting, and checkpoint-before-fail behavior.

^{Reviewed by Cursor Bugbot for commit e28208b. Bugbot is set up for automated code reviews on this repo. Configure here.}