Run commands and tests
Execute workspace commands and tests, then inspect exit codes, failures, and artifacts.
Try a task.NET runtime · MCP server
Run, test, and trace .NET
with your coding agent.
GliderTrace runs your commands and tests, stores the evidence as queryable sessions, and compares failures before and after a change.
dotnet tool exec "glider-trace@*" --no-http-cache --verbosity quiet -- --versionRequires the .NET 10 SDK and an MCP client.
tests / Orders.Tests
Query the stored session
Read the evidence, not the scrollback.
session · completed failure OrdersTests.Total_IncludesTax at OrderService.CalculateTotal Services/OrderService.cs:42
The agent queries the stored session instead of rerunning the command.
Illustrative workflow
FROM A FAILING RUN TO REVIEWABLE EVIDENCE
Give your agent the runtime facts.
Collect runtime evidence
Capture counters, traces, or explicitly acknowledged memory dumps, or import existing CI artifacts.
Explore the toolsCompare before and after
See fixed, new, and persisting failures between two sessions as evidence for your review.
When to use itKeep the agent you already use.
Connect GliderTrace through MCP. Choose your client for the installation guide.
Check project compatibility →Privacy
Your runs. Your machine.
GliderTrace runs commands and stores sessions on your computer. Your client controls how results reach its model.
Usage telemetry is enabled by default and can be disabled. Agent feedback can publish a report. Read the data policy and controls.
Privacy and telemetry →Pricing
Included with Glider.
Included with every Glider subscription. Free for non-commercial use and open source, with a 30-day evaluation at work.
Future packages will expire 90 days after release. Existing versions retain their original expiry. Keep the product updated to continue using it.
Pricing and license details →ALSO FROM GLIDERMCP
More tools for your codebase.
Connect GliderTrace to your next .NET bug.
Install the tool, connect your client, and run the failing command.