trace_run_tests

[solution]

Runs .NET tests and stores failures, output, and available TRX artifacts.

Before you use this tool

Executes test runners with your local permissions. Test runs can change workspace files and contact package sources. Check the workspace and runner policy. Supports dotnet test, dotnet vstest, and policy-validated native vstest.console. Use filters for expensive test suites.

Try it locally in the GliderTrace playground

Parameters

Name	Type	Required	Description
targetPath	string	No	Optional solution, project, or directory path under the workspace root. Defaults to the workspace root. Not valid with runner 'vstest' or 'vstest-console'.
configuration	string	No	Build configuration passed to dotnet test. Defaults to Release. Not valid with runner 'vstest' or 'vstest-console'.
filter	string	No	Optional dotnet test filter expression. With a VSTest runner, it is applied as TestCaseFilter.
verbosity	string	No	dotnet test verbosity: 'quiet', 'minimal', 'normal', 'detailed', or 'diagnostic'. Defaults to minimal. Not valid with runner 'vstest' or 'vstest-console'.
noRestore	boolean	No	Pass --no-restore to dotnet test. Defaults to false so fresh workspaces can restore normally. Default: `false`
collectBlame	boolean	No	Collect blame sequence artifacts when supported. Defaults to true. Default: `true`
rerunFailed	number	No	Number of times to rerun safely identified failed tests. Defaults to 0. A matching structured pass establishes a flaky candidate. Default: `0`
runner	string	No	Test runner: 'auto' (default, dotnet test with mode detection), 'dotnet-test', or 'vstest' (dotnet vstest against prebuilt assemblies). From GliderTrace 1.1.0: 'vstest-console' uses native VSTest and requires its executable policy. 'nunit-console' and 'msbuild-vstest' are reserved for a future release.
assemblies	json	No	Prebuilt test assembly paths under a configured workspace root. Requires runner 'vstest' or 'vstest-console'.
tests	json	No	Exact FullyQualifiedName test names, run through a VSTest TestCaseFilter so each name matches exactly. Requires runner 'vstest' or 'vstest-console'; cannot be combined with filter or testCaseFilter.
testCaseFilter	string	No	VSTest TestCaseFilter expression for trait/category selection. Requires runner 'vstest' or 'vstest-console'.
testAdapterPath	string	No	Directory containing test adapters, mapped to VSTest TestAdapterPath. Requires runner 'vstest' or 'vstest-console'.
platform	string	No	Target platform architecture for VSTest: 'x86', 'x64', or 'ARM'. Requires runner 'vstest' or 'vstest-console'.
framework	string	No	Target framework moniker for VSTest, for example Framework48. Requires runner 'vstest' or 'vstest-console'.
vstestConsolePath	string	No	Native VSTest executable path. Requires runner vstest-console.
runSettingsPath	string	No	Workspace-scoped VSTest settings. Collectors, custom loggers, and dump capture are rejected.
timeoutMs	number	No	Test timeout in milliseconds. A timeout stores a timeout status.
label	string	No	Optional short label for the stored session, for example nightly-baseline. Use it to find the session again with trace_list_sessions. Maximum 128 characters.

Examples

Replace sample paths and symbol keys with values from your workspace.

Run tests for the current workspace

Arguments

{}

Full MCP request

Full MCP request 
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "trace_run_tests",
    "arguments": {}
  }
}

Run a prebuilt legacy test assembly through vstest

Arguments 
{
  "runner": "vstest",
  "assemblies": [
    "Tests/bin/Debug/Sample.UnitTests.dll"
  ]
}

Full MCP request

Full MCP request 
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "trace_run_tests",
    "arguments": {
      "runner": "vstest",
      "assemblies": [
        "Tests/bin/Debug/Sample.UnitTests.dll"
      ]
    }
  }
}

Response Notes

Returns a test session id, result summary, TRX artifacts, stdout/stderr references, failure evidence, and rerun metadata. The summary adds knownFlakyFailures. flakyAnnotations carries the cross-session flake history for each failure the ledger has seen before. Each annotation lists failingSessions, flakyCandidateSessions, knownFlaky, firstSeen, and lastSeen. An annotation tags a failure; it never hides one.

From GliderTrace 1.0.0: Read testCountStatus: exact, partial, or unknown. testCountSource is trx or none; testCountReason explains uncertainty. testCountScope identifies the attempt. Counts remain numeric strings when known and use unknown otherwise. This also applies to initialFailedTests and finalFailedTests.

The reportedTotalTests, reportedPassedTests, reportedFailedTests, and reportedSkippedTests fields give available report subtotals. reportedTestCountScope is structured-reports. A report subtotal does not prove coverage of every requested target. An exit code of zero does not establish an exact test total.

Only a matching structured pass clears a failed test and establishes a flaky candidate. A skipped test or missing report leaves the failure unresolved. Capture coverage retains gaps from earlier attempts, even when a later attempt passes.

GliderTrace 1.0.0 returns compact JSON for every tool. Result fields appear beside success; there is no response-format parameter. success describes the tool call. Check the process or session status and exitCode separately. Scalar results and fields that conflict with reserved names appear under result.

To migrate an envelope consumer, read response.sessionId instead of response.data.sessionId. Stored sessions and exported files keep their formats. --max-response-chars limits response text to 60,000 UTF-16 characters by default. Use 0 to disable it, or at least 1,000. This limit counts text characters, not transport bytes or tokens.

outputLimit.truncated marks omitted response detail. It does not change capture completeness. Use the returned next tool and arguments when available. Do not repeat a command or feedback submission to recover omitted output. For event pages, keep the filters and cursor while reducing limit.

For session details, set includeArtifacts and includeTopFindings to false to reduce output. Check error, actionable meta, warnings, and coverage before drawing conclusions. Internal diagnostics stay in local server logs.