@vitest-agent/sdk 1.2.0 → 1.3.1

package/index.d.ts

@@ -1932,6 +1932,7 @@ type TestRun = typeof TestRun.Type;
  * @public
  */
 declare const TestHistory: Schema.Struct<{
+  modulePath: typeof Schema.String;
   fullName: typeof Schema.String;
   runs: Schema.Array$<Schema.Struct<{
     timestamp: typeof Schema.String;
@@ -1948,6 +1949,7 @@ declare const HistoryRecord: Schema.Struct<{
   project: typeof Schema.String;
   updatedAt: typeof Schema.String;
   tests: Schema.Array$<Schema.Struct<{
+    modulePath: typeof Schema.String;
     fullName: typeof Schema.String;
     runs: Schema.Array$<Schema.Struct<{
       timestamp: typeof Schema.String;
@@ -2596,7 +2598,7 @@ declare const DataStore_base: Context.TagClass<DataStore, "vitest-agent/DataStor
   readonly writeTestCases: (moduleId: number, tests: ReadonlyArray<TestCaseInput>) => Effect.Effect<ReadonlyArray<number>, DataStoreError>;
   readonly writeErrors: (runId: number, errors: ReadonlyArray<TestErrorInput>) => Effect.Effect<void, DataStoreError>;
   readonly writeCoverage: (runId: number, coverage: ReadonlyArray<FileCoverageInput>) => Effect.Effect<void, DataStoreError>;
-  readonly writeHistory: (project: string, fullName: string, runId: number, timestamp: string, state: string, duration: number | null, flaky: boolean, retryCount: number, errorMessage: string | null) => Effect.Effect<void, DataStoreError>;
+  readonly writeHistory: (project: string, fullName: string, modulePath: string, runId: number, timestamp: string, state: string, duration: number | null, flaky: boolean, retryCount: number, errorMessage: string | null) => Effect.Effect<void, DataStoreError>;
   readonly writeBaselines: (baselines: CoverageBaselines) => Effect.Effect<void, DataStoreError>;
   readonly writeTrends: (project: string, runId: number, entry: TrendEntry) => Effect.Effect<void, DataStoreError>;
   readonly writeSourceMap: (sourceFilePath: string, testModuleId: number, mappingType: string) => Effect.Effect<void, DataStoreError>;
@@ -2682,6 +2684,7 @@ interface ProjectRunSummary {
 /** @public */
 interface FlakyTest {
   readonly fullName: string;
+  readonly modulePath: string;
   readonly project: string;
   readonly passCount: number;
   readonly failCount: number;
@@ -2691,6 +2694,7 @@ interface FlakyTest {
 /** @public */
 interface PersistentFailure {
   readonly fullName: string;
+  readonly modulePath: string;
   readonly project: string;
   readonly consecutiveFailures: number;
   readonly firstFailedAt: string;
@@ -3208,9 +3212,22 @@ declare const FormatSelectorLive: Layer.Layer<FormatSelector, never, never>;
  * @public
  */
 interface TestOutcome {
+  readonly modulePath: string;
   readonly fullName: string;
   readonly state: "passed" | "failed";
 }
+/**
+ * Builds the composite (modulePath, fullName) key used to key the internal
+ * testMap and the returned classifications Map, so identically-named tests
+ * in different files classify independently instead of colliding.
+ *
+ * Uses `JSON.stringify` for an injective encoding: a plain delimiter such as
+ * a space is not collision-proof (e.g. `("a b", "c")` and `("a", "b c")` would
+ * both yield `"a b c"`), which would reintroduce the very collision class this
+ * key exists to prevent.
+ * @public
+ */
+declare const historyKey: (modulePath: string, fullName: string) => string;
 declare const HistoryTracker_base: Context.TagClass<HistoryTracker, "vitest-agent/HistoryTracker", {
   readonly classify: (project: string, testOutcomes: ReadonlyArray<TestOutcome>, timestamp: string) => Effect.Effect<{
     history: HistoryRecord;
@@ -4325,12 +4342,15 @@ interface ManifestRow {
  * @public
  */
 interface HistoryRow {
+  module_path: string;
   full_name: string;
   timestamp: string;
   state: string;
 }
 /**
- * A `Record<fullName, { runs: TestRun[] }>` keyed by test full name.
+ * A `Record<compositeKey, { modulePath, fullName, runs: TestRun[] }>` keyed by
+ * the injective `historyKey(modulePath, fullName)` string, so identically-named
+ * tests in different files are tracked as distinct entries.
  *
  * This is the internal representation returned by `assembleHistoryRecord`.
  * Callers that need the Effect Schema `HistoryRecord` type should adapt this
@@ -4338,7 +4358,9 @@ interface HistoryRow {
  * @public
  */
 interface AssembledHistoryRecord {
-  [fullName: string]: {
+  [compositeKey: string]: {
+    modulePath: string;
+    fullName: string;
     runs: TestRun[];
   };
 }
@@ -4351,10 +4373,13 @@ interface AssembledHistoryRecord {
  */
 declare function assembleManifest(rows: ManifestRow[], dbPath: string): CacheManifest;
 /**
- * Groups test_history rows by `full_name` into a keyed record.
+ * Groups test_history rows by the composite (module_path, full_name) key
+ * into a keyed record.
  *
- * Returns `Record<fullName, { runs: TestRun[] }>` where `runs` are the
- * accumulated test runs for that test, in the order they appear in `rows`.
+ * Returns `Record<compositeKey, { modulePath, fullName, runs: TestRun[] }>`
+ * where `runs` are the accumulated test runs for that test, in the order
+ * they appear in `rows`. Two rows sharing `full_name` but differing
+ * `module_path` are tracked as distinct entries rather than merged.
  *
  * Only rows with state `"passed"` or `"failed"` are included in runs
  * (matching the `TestRun` schema constraint).
@@ -5240,5 +5265,5 @@ declare function validateCoverageTargetsShape(input: unknown): CoverageTargetsSh
  */
 declare const CURRENT_SDK_VERSION: string;
 //#endregion
-export { AcceptanceMetrics, ActionSeverity, ActorType, Agent, AgentConsoleMode, AgentContext, AgentId, AgentNotFoundError, AgentPluginOptions, AgentReport, AgentReporterOptions, AgentTreeNode, AnsiOptions, ArtifactKind, AssembledHistoryRecord, AssociateRunSessionInput, BehaviorAbandonedEvent, BehaviorAddedEvent, BehaviorCompletedEvent, BehaviorDetail, BehaviorNotFoundError, BehaviorRow, BehaviorScopedEventType, BehaviorScopedEventTypes, BehaviorStartedEvent, BehaviorStatus, BehaviorsReadyEvent, BlockedEvent, CURRENT_SDK_VERSION, CacheManifest, CacheManifestEntry, CandidateContext, CellOptions, ChangeKind, ChannelEvent, ChatId, CiConsoleMode, CitedArtifact, CitedArtifactRow, CommitChangesEntry, ConfigLive, ConsoleFormatOptions, ConsoleLeakEntry, ConsoleLeakFile, ConsoleLeakTask, ConsoleLeaks, ConsoleLogLevel, ConsoleMode, ConsoleOutputMode, ConsoleOutputs, ConversationId, ConversationNotFoundError, CoverageBaselines, FileCoverageReport as CoverageFile, FileCoverageReport, CoverageGap, CoverageInput, CoverageLevel, CoverageLevelName, CoverageMetric, CoverageRenderState, CoverageReport, CoverageTargets, CoverageTargetsMetrics, CoverageTargetsShapeError, CoverageTargetsShapeInfo, CoverageTargetsShapeResult, CoverageTargetsShapeWarning, CoverageTotals, CreateBehaviorInput, CreateGoalInput, CurrentTddPhase, DATABASE_FILENAME, DataReader, DataReaderLive, DataStore, DataStoreError, DataStoreLive, DenialReason, DetailLevel, DetailResolver, DetailResolverLive, DiscoveryError, DiscoveryRegistry, DiscoveryRegistryLive, DispatchInputs, EndTddTaskInput, Environment, EnvironmentDetector, EnvironmentDetectorLive, EnvironmentDetectorTest, Executor, ExecutorResolver, ExecutorResolverLive, FailureRecord, FailureSignatureDetail, FailureSignatureInput, FailureSignatureWriteInput, FileCoverageInput, FileEditPayload, FileSystemAdapter, FlakyTest, FormatSelector, FormatSelectorLive, type FormatTriageOptions, type FormatWrapupOptions, Formatter, FormatterContext, FunctionBoundary, GfmFormatter, GitCommandError, GoalAbandonedEvent, GoalAddedEvent, GoalCompletedEvent, GoalDetail, GoalNotFoundError, GoalRow, GoalStartedEvent, GoalStatus, GoalsReadyEvent, HistoryRecord, HistoryRow, HistoryTracker, HistoryTrackerLive, HistoryTrackerTest, HistoryTrackerTestState, HookFirePayload, HookStatus, HookType, HostKind, HostMetadataParseError, HostMetadataResult, HumanConsoleMode, HypothesisDetail, HypothesisInput, HypothesisPayload, IDEMPOTENCY_ROOT_SENTINEL, IdempotencyHit, IdempotencyInput, IdempotentResponseInput, IllegalStatusTransitionEntity, IllegalStatusTransitionError, JsonFormatter, KnownProject, KnownProjectInput, LoggerLive, ManifestRow, MapSessionInput, MapSessionOutput, MarkdownFormatter, MetricThresholds, ModuleInput, ModuleListEntry, ModuleRecord, ModuleReport, NoteInput, NotePayload, NoteRow, Osc8Options, OutputFormat, OutputPipelineLive, OutputRenderer, OutputRendererLive, PackageManager, PathResolutionError, PathResolutionLive, PatternThresholds, PerClientSessionMapReader, PerClientSessionMapReaderLive, PerClientSessionMapWriter, PerClientSessionMapWriterLive, PersistentFailure, Phase, PhaseTransitionContext, PhaseTransitionEvent, PhaseTransitionResult, ProbeMissError, ProjectDiscovery, ProjectDiscoveryLive, ProjectDiscoveryTest, ProjectIdentity, ProjectIdentityCandidate, ProjectIdentityCandidates, ProjectIdentityLive, ProjectIdentityNotResolvableError, ProjectIdentitySource, ProjectIdentityTest, ProjectKey, ProjectRunSummary, ProjectSummary, RegisterAgentInput, RegisterAgentResult, RegistrationConflictError, Remediation, RenderState, RenderTotals, RenderedOutput, ReportError, ReportSummary, ReporterKit, ReporterRenderInput, ResolveDataPathOptions, ResolveProjectIdentityOptions, ResolvedIdentity, ResolvedReporterConfig, ResolvedThresholds, RunChangedFile, RunContext, RunContext$, RunContextLive, RunContextService, RunContextTest, RunEvent, RunEventByTag, RunHealth, RunInvocationMethod, RunOutcome, RunShape, SessionCompleteEvent, SessionContextRow, SessionDetail, SessionInput, SessionNotFoundError, SessionOutcome, SettingsInput, SettingsListEntry, SettingsRow, SidecarTimeoutError, SilentFormatter, StackFrameInput, SuggestedActionRecord, SuiteInput, SuiteListEntry, TagCountEntry, TagInventoryRow, TddArtifactDetail, TddArtifactRow, TddPhaseDetail, TddTaskAlreadyEndedError, TddTaskDetail, TddTaskEndOutcome, TddTaskId, TddTaskInput, TddTaskNotFoundError, TddTaskSummary, TerminalFormatter, TestCaseInput, TestClassification, TestError, TestErrorInput, TestFileEntry, TestHistory, TestListEntry, TestOutcome, TestRecord, TestReport, TestRun, TestRunInput, TestRunReason, TestState, ToolCallPayload, ToolResultPayload, Transport, TrendEntry, TrendRecord, TrendSummary, TurnInput, TurnPayload, TurnSearchOptions, TurnSummary, UpdateBehaviorInput, UpdateGoalInput, UserPromptPayload, ValidateHypothesisInput, VitestAgentConfig, VitestAgentConfigFile, VitestAgentConfigFileService, VitestAgentReporter, VitestAgentReporterFactory, VitestModuleDiagnostic, VitestModuleError, VitestParsedStack, VitestTestCase, VitestTestDiagnostic, VitestTestError, VitestTestModule, VitestTestResult, VitestTestSuite, type WrapupKind, WriteCommitInput, WriteRunChangedFilesInput, WriteTddArtifactInput, WriteTddPhaseInput, WriteTddPhaseOutput, _resetMigrationCacheForTesting, ansi, assembleHistoryRecord, assembleManifest, buildAgentReport, buildConsoleLeaks, buildEnvPrefix, canonicalizeGitUrl, ciAnnotationsFormatter, classifyTest, collectConsoleLeakEntries, compressLines, computeFailureSignature, computeTrend, deriveIdempotencyKey, detectPackageManager, detectVitestScripts, determineTier, ensureMigrated, extractSqlReason, findFunctionBoundary, formatConsoleMarkdown, formatFatalError, formatGfm, formatTriageEffect, formatWrapupEffect, getRecentDirection, getRunCommand, getWorstMetric, gitUrlToProjectKey, hashTargets, initialRenderState, isTimeoutError, isVitestInvocation, migration as migration0001, normalizeAssertionShape, normalizeWorkspaceKey, osc8, probeHostMetadataFromEnv, migration$1 as registryMigration0001, relativePath, requiredArtifactForTransition, resolveCoverageInput, resolveDataPath, resolveLogFile, resolveLogLevel, resolveProjectIdentityFromCandidates, resolveProjectKeyFromCwd, resolveWorkspaceKey, rewriteBashCommand, safeFilename, migration$2 as sessionMapMigration0001, stripAnsi, transitionEnforcesBehaviorMatch, validateCoverageConfig, validateCoverageTargetsShape, validatePhaseTransition };
+export { AcceptanceMetrics, ActionSeverity, ActorType, Agent, AgentConsoleMode, AgentContext, AgentId, AgentNotFoundError, AgentPluginOptions, AgentReport, AgentReporterOptions, AgentTreeNode, AnsiOptions, ArtifactKind, AssembledHistoryRecord, AssociateRunSessionInput, BehaviorAbandonedEvent, BehaviorAddedEvent, BehaviorCompletedEvent, BehaviorDetail, BehaviorNotFoundError, BehaviorRow, BehaviorScopedEventType, BehaviorScopedEventTypes, BehaviorStartedEvent, BehaviorStatus, BehaviorsReadyEvent, BlockedEvent, CURRENT_SDK_VERSION, CacheManifest, CacheManifestEntry, CandidateContext, CellOptions, ChangeKind, ChannelEvent, ChatId, CiConsoleMode, CitedArtifact, CitedArtifactRow, CommitChangesEntry, ConfigLive, ConsoleFormatOptions, ConsoleLeakEntry, ConsoleLeakFile, ConsoleLeakTask, ConsoleLeaks, ConsoleLogLevel, ConsoleMode, ConsoleOutputMode, ConsoleOutputs, ConversationId, ConversationNotFoundError, CoverageBaselines, FileCoverageReport as CoverageFile, FileCoverageReport, CoverageGap, CoverageInput, CoverageLevel, CoverageLevelName, CoverageMetric, CoverageRenderState, CoverageReport, CoverageTargets, CoverageTargetsMetrics, CoverageTargetsShapeError, CoverageTargetsShapeInfo, CoverageTargetsShapeResult, CoverageTargetsShapeWarning, CoverageTotals, CreateBehaviorInput, CreateGoalInput, CurrentTddPhase, DATABASE_FILENAME, DataReader, DataReaderLive, DataStore, DataStoreError, DataStoreLive, DenialReason, DetailLevel, DetailResolver, DetailResolverLive, DiscoveryError, DiscoveryRegistry, DiscoveryRegistryLive, DispatchInputs, EndTddTaskInput, Environment, EnvironmentDetector, EnvironmentDetectorLive, EnvironmentDetectorTest, Executor, ExecutorResolver, ExecutorResolverLive, FailureRecord, FailureSignatureDetail, FailureSignatureInput, FailureSignatureWriteInput, FileCoverageInput, FileEditPayload, FileSystemAdapter, FlakyTest, FormatSelector, FormatSelectorLive, type FormatTriageOptions, type FormatWrapupOptions, Formatter, FormatterContext, FunctionBoundary, GfmFormatter, GitCommandError, GoalAbandonedEvent, GoalAddedEvent, GoalCompletedEvent, GoalDetail, GoalNotFoundError, GoalRow, GoalStartedEvent, GoalStatus, GoalsReadyEvent, HistoryRecord, HistoryRow, HistoryTracker, HistoryTrackerLive, HistoryTrackerTest, HistoryTrackerTestState, HookFirePayload, HookStatus, HookType, HostKind, HostMetadataParseError, HostMetadataResult, HumanConsoleMode, HypothesisDetail, HypothesisInput, HypothesisPayload, IDEMPOTENCY_ROOT_SENTINEL, IdempotencyHit, IdempotencyInput, IdempotentResponseInput, IllegalStatusTransitionEntity, IllegalStatusTransitionError, JsonFormatter, KnownProject, KnownProjectInput, LoggerLive, ManifestRow, MapSessionInput, MapSessionOutput, MarkdownFormatter, MetricThresholds, ModuleInput, ModuleListEntry, ModuleRecord, ModuleReport, NoteInput, NotePayload, NoteRow, Osc8Options, OutputFormat, OutputPipelineLive, OutputRenderer, OutputRendererLive, PackageManager, PathResolutionError, PathResolutionLive, PatternThresholds, PerClientSessionMapReader, PerClientSessionMapReaderLive, PerClientSessionMapWriter, PerClientSessionMapWriterLive, PersistentFailure, Phase, PhaseTransitionContext, PhaseTransitionEvent, PhaseTransitionResult, ProbeMissError, ProjectDiscovery, ProjectDiscoveryLive, ProjectDiscoveryTest, ProjectIdentity, ProjectIdentityCandidate, ProjectIdentityCandidates, ProjectIdentityLive, ProjectIdentityNotResolvableError, ProjectIdentitySource, ProjectIdentityTest, ProjectKey, ProjectRunSummary, ProjectSummary, RegisterAgentInput, RegisterAgentResult, RegistrationConflictError, Remediation, RenderState, RenderTotals, RenderedOutput, ReportError, ReportSummary, ReporterKit, ReporterRenderInput, ResolveDataPathOptions, ResolveProjectIdentityOptions, ResolvedIdentity, ResolvedReporterConfig, ResolvedThresholds, RunChangedFile, RunContext, RunContext$, RunContextLive, RunContextService, RunContextTest, RunEvent, RunEventByTag, RunHealth, RunInvocationMethod, RunOutcome, RunShape, SessionCompleteEvent, SessionContextRow, SessionDetail, SessionInput, SessionNotFoundError, SessionOutcome, SettingsInput, SettingsListEntry, SettingsRow, SidecarTimeoutError, SilentFormatter, StackFrameInput, SuggestedActionRecord, SuiteInput, SuiteListEntry, TagCountEntry, TagInventoryRow, TddArtifactDetail, TddArtifactRow, TddPhaseDetail, TddTaskAlreadyEndedError, TddTaskDetail, TddTaskEndOutcome, TddTaskId, TddTaskInput, TddTaskNotFoundError, TddTaskSummary, TerminalFormatter, TestCaseInput, TestClassification, TestError, TestErrorInput, TestFileEntry, TestHistory, TestListEntry, TestOutcome, TestRecord, TestReport, TestRun, TestRunInput, TestRunReason, TestState, ToolCallPayload, ToolResultPayload, Transport, TrendEntry, TrendRecord, TrendSummary, TurnInput, TurnPayload, TurnSearchOptions, TurnSummary, UpdateBehaviorInput, UpdateGoalInput, UserPromptPayload, ValidateHypothesisInput, VitestAgentConfig, VitestAgentConfigFile, VitestAgentConfigFileService, VitestAgentReporter, VitestAgentReporterFactory, VitestModuleDiagnostic, VitestModuleError, VitestParsedStack, VitestTestCase, VitestTestDiagnostic, VitestTestError, VitestTestModule, VitestTestResult, VitestTestSuite, type WrapupKind, WriteCommitInput, WriteRunChangedFilesInput, WriteTddArtifactInput, WriteTddPhaseInput, WriteTddPhaseOutput, _resetMigrationCacheForTesting, ansi, assembleHistoryRecord, assembleManifest, buildAgentReport, buildConsoleLeaks, buildEnvPrefix, canonicalizeGitUrl, ciAnnotationsFormatter, classifyTest, collectConsoleLeakEntries, compressLines, computeFailureSignature, computeTrend, deriveIdempotencyKey, detectPackageManager, detectVitestScripts, determineTier, ensureMigrated, extractSqlReason, findFunctionBoundary, formatConsoleMarkdown, formatFatalError, formatGfm, formatTriageEffect, formatWrapupEffect, getRecentDirection, getRunCommand, getWorstMetric, gitUrlToProjectKey, hashTargets, historyKey, initialRenderState, isTimeoutError, isVitestInvocation, migration as migration0001, normalizeAssertionShape, normalizeWorkspaceKey, osc8, probeHostMetadataFromEnv, migration$1 as registryMigration0001, relativePath, requiredArtifactForTransition, resolveCoverageInput, resolveDataPath, resolveLogFile, resolveLogLevel, resolveProjectIdentityFromCandidates, resolveProjectKeyFromCwd, resolveWorkspaceKey, rewriteBashCommand, safeFilename, migration$2 as sessionMapMigration0001, stripAnsi, transitionEnforcesBehaviorMatch, validateCoverageConfig, validateCoverageTargetsShape, validatePhaseTransition };
 //# sourceMappingURL=index.d.ts.map

package/index.js

@@ -20,6 +20,7 @@ import { VitestAgentConfig } from "./schemas/Config.js";
 import { VitestAgentConfigFile } from "./services/Config.js";
 import { ConfigLive } from "./layers/ConfigLive.js";
 import { DataReader } from "./services/DataReader.js";
+import { HistoryTracker, historyKey } from "./services/HistoryTracker.js";
 import { DataReaderLive } from "./layers/DataReaderLive.js";
 import { ActorType, AgentId, ChatId, ConversationId, HostKind, ProjectKey, TddTaskId } from "./schemas/Identity.js";
 import { Agent, IdempotencyHit } from "./schemas/Agent.js";
@@ -36,7 +37,6 @@ import { ExecutorResolver } from "./services/ExecutorResolver.js";
 import { ExecutorResolverLive } from "./layers/ExecutorResolverLive.js";
 import { FormatSelector } from "./services/FormatSelector.js";
 import { FormatSelectorLive } from "./layers/FormatSelectorLive.js";
-import { HistoryTracker } from "./services/HistoryTracker.js";
 import { classifyTest } from "./utils/classify-test.js";
 import { HistoryTrackerLive } from "./layers/HistoryTrackerLive.js";
 import { HistoryTrackerTest } from "./layers/HistoryTrackerTest.js";
@@ -115,7 +115,7 @@ import { requiredArtifactForTransition, transitionEnforcesBehaviorMatch, validat
 * substituted yet. Exported for version introspection by downstream tooling.
 * @public
 */
-const CURRENT_SDK_VERSION = "1.2.0";
+const CURRENT_SDK_VERSION = "1.3.1";
 
 //#endregion
-export { ActionSeverity, ActorType, Agent, AgentConsoleMode, AgentContext, AgentId, AgentNotFoundError, AgentPluginOptions, AgentReport, AgentReporterOptions, BehaviorAbandonedEvent, BehaviorAddedEvent, BehaviorCompletedEvent, BehaviorDetail, BehaviorNotFoundError, BehaviorRow, BehaviorScopedEventTypes, BehaviorStartedEvent, BehaviorStatus, BehaviorsReadyEvent, BlockedEvent, CURRENT_SDK_VERSION, CacheManifest, CacheManifestEntry, ChannelEvent, ChatId, CiConsoleMode, ConfigLive, ConsoleLeakFile, ConsoleLeaks, ConsoleLogLevel, ConsoleMode, ConsoleOutputMode, ConsoleOutputs, ConversationId, ConversationNotFoundError, CoverageBaselines, FileCoverageReport as CoverageFile, CoverageGap, CoverageLevel, CoverageMetric, CoverageRenderState, CoverageReport, CoverageTargets, CoverageTargetsMetrics, CoverageTotals, DATABASE_FILENAME, DataReader, DataReaderLive, DataStore, DataStoreError, DataStoreLive, DetailLevel, DetailResolver, DetailResolverLive, DiscoveryError, DiscoveryRegistry, DiscoveryRegistryLive, Environment, EnvironmentDetector, EnvironmentDetectorLive, EnvironmentDetectorTest, Executor, ExecutorResolver, ExecutorResolverLive, FailureRecord, FileCoverageReport, FileEditPayload, FormatSelector, FormatSelectorLive, GfmFormatter, GitCommandError, GoalAbandonedEvent, GoalAddedEvent, GoalCompletedEvent, GoalDetail, GoalNotFoundError, GoalRow, GoalStartedEvent, GoalStatus, GoalsReadyEvent, HistoryRecord, HistoryTracker, HistoryTrackerLive, HistoryTrackerTest, HookFirePayload, HookStatus, HookType, HostKind, HostMetadataParseError, HumanConsoleMode, HypothesisPayload, IDEMPOTENCY_ROOT_SENTINEL, IdempotencyHit, IllegalStatusTransitionError, JsonFormatter, LoggerLive, MarkdownFormatter, MetricThresholds, ModuleRecord, ModuleReport, NotePayload, OutputFormat, OutputPipelineLive, OutputRenderer, OutputRendererLive, PackageManager, PathResolutionError, PathResolutionLive, PatternThresholds, PerClientSessionMapReader, PerClientSessionMapReaderLive, PerClientSessionMapWriter, PerClientSessionMapWriterLive, PhaseTransitionEvent, ProbeMissError, ProjectDiscovery, ProjectDiscoveryLive, ProjectDiscoveryTest, ProjectIdentity, ProjectIdentityLive, ProjectIdentityNotResolvableError, ProjectIdentityTest, ProjectKey, RegistrationConflictError, RenderState, RenderTotals, ReportError, ReportSummary, ResolvedThresholds, RunContext, RunContext$, RunContextLive, RunContextService, RunContextTest, RunEvent, SessionCompleteEvent, SessionNotFoundError, SessionOutcome, SidecarTimeoutError, SilentFormatter, SuggestedActionRecord, TagCountEntry, TddTaskAlreadyEndedError, TddTaskId, TddTaskNotFoundError, TerminalFormatter, TestClassification, TestHistory, TestRecord, TestReport, TestRun, TestRunReason, TestState, ToolCallPayload, ToolResultPayload, Transport, TrendEntry, TrendRecord, TurnPayload, UserPromptPayload, VitestAgentConfig, VitestAgentConfigFile, _resetMigrationCacheForTesting, ansi, assembleHistoryRecord, assembleManifest, buildAgentReport, buildConsoleLeaks, buildEnvPrefix, canonicalizeGitUrl, ciAnnotationsFormatter, classifyTest, collectConsoleLeakEntries, compressLines, computeFailureSignature, computeTrend, deriveIdempotencyKey, detectPackageManager, detectVitestScripts, determineTier, ensureMigrated, extractSqlReason, findFunctionBoundary, formatConsoleMarkdown, formatFatalError, formatGfm, formatTriageEffect, formatWrapupEffect, getRecentDirection, getRunCommand, getWorstMetric, gitUrlToProjectKey, hashTargets, initialRenderState, isTimeoutError, isVitestInvocation, migration as migration0001, normalizeAssertionShape, normalizeWorkspaceKey, osc8, probeHostMetadataFromEnv, migration$1 as registryMigration0001, relativePath, requiredArtifactForTransition, resolveCoverageInput, resolveDataPath, resolveLogFile, resolveLogLevel, resolveProjectIdentityFromCandidates, resolveProjectKeyFromCwd, resolveWorkspaceKey, rewriteBashCommand, safeFilename, migration$2 as sessionMapMigration0001, stripAnsi, transitionEnforcesBehaviorMatch, validateCoverageConfig, validateCoverageTargetsShape, validatePhaseTransition };
+export { ActionSeverity, ActorType, Agent, AgentConsoleMode, AgentContext, AgentId, AgentNotFoundError, AgentPluginOptions, AgentReport, AgentReporterOptions, BehaviorAbandonedEvent, BehaviorAddedEvent, BehaviorCompletedEvent, BehaviorDetail, BehaviorNotFoundError, BehaviorRow, BehaviorScopedEventTypes, BehaviorStartedEvent, BehaviorStatus, BehaviorsReadyEvent, BlockedEvent, CURRENT_SDK_VERSION, CacheManifest, CacheManifestEntry, ChannelEvent, ChatId, CiConsoleMode, ConfigLive, ConsoleLeakFile, ConsoleLeaks, ConsoleLogLevel, ConsoleMode, ConsoleOutputMode, ConsoleOutputs, ConversationId, ConversationNotFoundError, CoverageBaselines, FileCoverageReport as CoverageFile, CoverageGap, CoverageLevel, CoverageMetric, CoverageRenderState, CoverageReport, CoverageTargets, CoverageTargetsMetrics, CoverageTotals, DATABASE_FILENAME, DataReader, DataReaderLive, DataStore, DataStoreError, DataStoreLive, DetailLevel, DetailResolver, DetailResolverLive, DiscoveryError, DiscoveryRegistry, DiscoveryRegistryLive, Environment, EnvironmentDetector, EnvironmentDetectorLive, EnvironmentDetectorTest, Executor, ExecutorResolver, ExecutorResolverLive, FailureRecord, FileCoverageReport, FileEditPayload, FormatSelector, FormatSelectorLive, GfmFormatter, GitCommandError, GoalAbandonedEvent, GoalAddedEvent, GoalCompletedEvent, GoalDetail, GoalNotFoundError, GoalRow, GoalStartedEvent, GoalStatus, GoalsReadyEvent, HistoryRecord, HistoryTracker, HistoryTrackerLive, HistoryTrackerTest, HookFirePayload, HookStatus, HookType, HostKind, HostMetadataParseError, HumanConsoleMode, HypothesisPayload, IDEMPOTENCY_ROOT_SENTINEL, IdempotencyHit, IllegalStatusTransitionError, JsonFormatter, LoggerLive, MarkdownFormatter, MetricThresholds, ModuleRecord, ModuleReport, NotePayload, OutputFormat, OutputPipelineLive, OutputRenderer, OutputRendererLive, PackageManager, PathResolutionError, PathResolutionLive, PatternThresholds, PerClientSessionMapReader, PerClientSessionMapReaderLive, PerClientSessionMapWriter, PerClientSessionMapWriterLive, PhaseTransitionEvent, ProbeMissError, ProjectDiscovery, ProjectDiscoveryLive, ProjectDiscoveryTest, ProjectIdentity, ProjectIdentityLive, ProjectIdentityNotResolvableError, ProjectIdentityTest, ProjectKey, RegistrationConflictError, RenderState, RenderTotals, ReportError, ReportSummary, ResolvedThresholds, RunContext, RunContext$, RunContextLive, RunContextService, RunContextTest, RunEvent, SessionCompleteEvent, SessionNotFoundError, SessionOutcome, SidecarTimeoutError, SilentFormatter, SuggestedActionRecord, TagCountEntry, TddTaskAlreadyEndedError, TddTaskId, TddTaskNotFoundError, TerminalFormatter, TestClassification, TestHistory, TestRecord, TestReport, TestRun, TestRunReason, TestState, ToolCallPayload, ToolResultPayload, Transport, TrendEntry, TrendRecord, TurnPayload, UserPromptPayload, VitestAgentConfig, VitestAgentConfigFile, _resetMigrationCacheForTesting, ansi, assembleHistoryRecord, assembleManifest, buildAgentReport, buildConsoleLeaks, buildEnvPrefix, canonicalizeGitUrl, ciAnnotationsFormatter, classifyTest, collectConsoleLeakEntries, compressLines, computeFailureSignature, computeTrend, deriveIdempotencyKey, detectPackageManager, detectVitestScripts, determineTier, ensureMigrated, extractSqlReason, findFunctionBoundary, formatConsoleMarkdown, formatFatalError, formatGfm, formatTriageEffect, formatWrapupEffect, getRecentDirection, getRunCommand, getWorstMetric, gitUrlToProjectKey, hashTargets, historyKey, initialRenderState, isTimeoutError, isVitestInvocation, migration as migration0001, normalizeAssertionShape, normalizeWorkspaceKey, osc8, probeHostMetadataFromEnv, migration$1 as registryMigration0001, relativePath, requiredArtifactForTransition, resolveCoverageInput, resolveDataPath, resolveLogFile, resolveLogLevel, resolveProjectIdentityFromCandidates, resolveProjectKeyFromCwd, resolveWorkspaceKey, rewriteBashCommand, safeFilename, migration$2 as sessionMapMigration0001, stripAnsi, transitionEnforcesBehaviorMatch, validateCoverageConfig, validateCoverageTargetsShape, validatePhaseTransition };

package/layers/DataReaderLive.js

@@ -1,5 +1,6 @@
 import { DataStoreError, extractSqlReason } from "../errors/DataStoreError.js";
 import { DataReader } from "../services/DataReader.js";
+import { historyKey } from "../services/HistoryTracker.js";
 import { Effect, Layer, Option } from "effect";
 import { SqlClient } from "@effect/sql/SqlClient";
 
@@ -195,24 +196,30 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 	})));
 	const getHistory = (project) => Effect.gen(function* () {
 		yield* Effect.logDebug("getHistory").pipe(Effect.annotateLogs({ project }));
-		const rows = yield* sql`SELECT full_name, timestamp, state
+		const rows = yield* sql`SELECT module_path, full_name, timestamp, state
 					FROM test_history
 					WHERE project = ${project}
-					ORDER BY full_name, timestamp DESC`;
+					ORDER BY module_path, full_name, timestamp DESC`;
 		const testsMap = /* @__PURE__ */ new Map();
 		for (const row of rows) {
 			if (row.state !== "passed" && row.state !== "failed") continue;
-			const existing = testsMap.get(row.full_name);
-			if (existing) existing.push({
+			const key = historyKey(row.module_path, row.full_name);
+			const existing = testsMap.get(key);
+			if (existing) existing.runs.push({
 				timestamp: row.timestamp,
 				state: row.state
 			});
-			else testsMap.set(row.full_name, [{
-				timestamp: row.timestamp,
-				state: row.state
-			}]);
+			else testsMap.set(key, {
+				modulePath: row.module_path,
+				fullName: row.full_name,
+				runs: [{
+					timestamp: row.timestamp,
+					state: row.state
+				}]
+			});
 		}
-		const tests = Array.from(testsMap.entries()).map(([fullName, runs]) => ({
+		const tests = Array.from(testsMap.values()).map(({ modulePath, fullName, runs }) => ({
+			modulePath,
 			fullName,
 			runs
 		}));
@@ -307,27 +314,31 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 		yield* Effect.logDebug("getFlaky").pipe(Effect.annotateLogs({ project }));
 		return (yield* sql`SELECT
 						th1.full_name,
+						th1.module_path,
 						th1.project,
 						SUM(CASE WHEN th1.state = 'passed' THEN 1 ELSE 0 END) as pass_count,
 						SUM(CASE WHEN th1.state = 'failed' THEN 1 ELSE 0 END) as fail_count,
 						(SELECT state FROM test_history th2
 						 WHERE th2.full_name = th1.full_name
+						   AND th2.module_path = th1.module_path
 						   AND th2.project = th1.project
 						 ORDER BY timestamp DESC LIMIT 1) as last_state,
 						MAX(th1.timestamp) as last_timestamp
 					FROM test_history th1
 					WHERE th1.project = ${project}
-					GROUP BY th1.full_name, th1.project
+					GROUP BY th1.full_name, th1.module_path, th1.project
 					-- Flaky means oscillation, not a clean recovery. A test with both
 						-- passes and fails is only flaky when at least one failure occurs at
 						-- or after the earliest pass (a fail-after-pass regression). A
 						-- monotonic red->green cycle — every failure precedes every pass — is
 						-- a recovery, not flakiness, so it is excluded here. ISO-8601
-						-- timestamps compare lexicographically.
+						-- timestamps compare lexicographically. Grouping keys on module_path
+						-- so same-named tests in different files stay distinct series.
 						HAVING pass_count > 0 AND fail_count > 0
 							AND MAX(CASE WHEN th1.state = 'failed' THEN th1.timestamp END)
 								>= MIN(CASE WHEN th1.state = 'passed' THEN th1.timestamp END)`).map((r) => ({
 			fullName: r.full_name,
+			modulePath: r.module_path,
 			project: r.project,
 			passCount: r.pass_count,
 			failCount: r.fail_count,
@@ -342,36 +353,38 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 	const getPersistentFailures = (project) => Effect.gen(function* () {
 		yield* Effect.logDebug("getPersistentFailures").pipe(Effect.annotateLogs({ project }));
 		return (yield* sql`WITH ranked AS (
-					SELECT full_name, project, state, timestamp, error_message,
+					SELECT full_name, module_path, project, state, timestamp, error_message,
 						ROW_NUMBER() OVER (
-							PARTITION BY project, full_name
+							PARTITION BY project, module_path, full_name
 							ORDER BY timestamp DESC
 						) as rn
 					FROM test_history
 					WHERE project = ${project}
 				),
 				streak AS (
-					SELECT full_name, project, state, timestamp, error_message, rn,
+					SELECT full_name, module_path, project, state, timestamp, error_message, rn,
 						MIN(CASE WHEN state != 'failed' THEN rn END) OVER (
-							PARTITION BY project, full_name
+							PARTITION BY project, module_path, full_name
 						) as first_pass_rn
 					FROM ranked
 				)
 				SELECT
-					full_name, project,
+					full_name, module_path, project,
 					COUNT(*) as consecutive_failures,
 					MIN(timestamp) as first_failed_at,
 					MAX(timestamp) as last_failed_at,
 					(SELECT error_message FROM streak s2
 					 WHERE s2.full_name = streak.full_name
+					   AND s2.module_path = streak.module_path
 					   AND s2.project = streak.project
 					   AND s2.rn = 1) as last_error_message
 				FROM streak
 				WHERE state = 'failed'
 					AND (first_pass_rn IS NULL OR rn < first_pass_rn)
-				GROUP BY full_name, project
+				GROUP BY full_name, module_path, project
 				HAVING consecutive_failures >= 2`).map((row) => ({
 			fullName: row.full_name,
+			modulePath: row.module_path,
 			project: row.project,
 			consecutiveFailures: row.consecutive_failures,
 			firstFailedAt: row.first_failed_at,
@@ -409,9 +422,10 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 					WHERE project = ${project}
 					ORDER BY timestamp DESC LIMIT 1`;
 		if (runs.length === 0) return Option.none();
+		const runId = runs[0].id;
 		const fileCoverageRows = yield* sql`SELECT f.path as file_path, fc.statements, fc.branches, fc.functions, fc.lines, fc.uncovered_lines
 					FROM file_coverage fc JOIN files f ON f.id = fc.file_id
-					WHERE fc.run_id = ${runs[0].id}`;
+					WHERE fc.run_id = ${runId}`;
 		const trendRows = yield* sql`SELECT statements, branches, functions, lines
 					FROM coverage_trends
 					WHERE project = ${project}
@@ -674,11 +688,12 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 					WHERE project = ${project}
 					ORDER BY timestamp DESC LIMIT 1`;
 		if (runs.length === 0) return Option.none();
+		const runId = runs[0].id;
 		const rows = yield* sql`SELECT tc.id, tc.full_name, tc.state, tc.duration, f.path as relative_module_id, tc.classification
 					FROM test_cases tc
 					JOIN test_modules tm ON tm.id = tc.module_id
 					JOIN files f ON f.id = tm.file_id
-					WHERE tm.run_id = ${runs[0].id} AND tc.full_name = ${fullName}
+					WHERE tm.run_id = ${runId} AND tc.full_name = ${fullName}
 					LIMIT 1`;
 		if (rows.length === 0) return Option.none();
 		const r = rows[0];
@@ -744,12 +759,13 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 					WHERE project = ${project}
 					ORDER BY timestamp DESC LIMIT 1`;
 		if (runs.length === 0) return [];
+		const runId = runs[0].id;
 		return (yield* sql`SELECT tm.id, f.path as file_path, tm.state,
 						(SELECT COUNT(*) FROM test_cases tc WHERE tc.module_id = tm.id) as test_count,
 						tm.duration
 					FROM test_modules tm
 					JOIN files f ON f.id = tm.file_id
-					WHERE tm.run_id = ${runs[0].id}
+					WHERE tm.run_id = ${runId}
 					ORDER BY f.path`).map((r) => ({
 			id: r.id,
 			file: r.file_path,
@@ -847,7 +863,8 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 	})));
 	const findSessionsByChatPrefix = (prefix) => Effect.gen(function* () {
 		yield* Effect.logDebug("findSessionsByChatPrefix").pipe(Effect.annotateLogs({ prefix }));
-		return (yield* sql`SELECT id, chat_id, project, cwd, agent_kind, agent_type, parent_session_id, triage_was_non_empty, started_at, ended_at, end_reason FROM sessions WHERE chat_id LIKE ${`${prefix}%`} ESCAPE '\\' ORDER BY started_at DESC`).map(sessionRowToDetail);
+		const pattern = `${prefix}%`;
+		return (yield* sql`SELECT id, chat_id, project, cwd, agent_kind, agent_type, parent_session_id, triage_was_non_empty, started_at, ended_at, end_reason FROM sessions WHERE chat_id LIKE ${pattern} ESCAPE '\\' ORDER BY started_at DESC`).map(sessionRowToDetail);
 	}).pipe(Effect.annotateLogs("service", "DataReader"), Effect.mapError((e) => new DataStoreError({
 		operation: "read",
 		table: "sessions",
@@ -866,7 +883,10 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 	const searchTurns = (options) => Effect.gen(function* () {
 		yield* Effect.logDebug("searchTurns").pipe(Effect.annotateLogs({ ...options }));
 		const limit = options.limit ?? 100;
-		return (yield* sql`SELECT id, session_id, turn_no, type, payload, occurred_at FROM turns WHERE 1=1${options.sessionId !== void 0 ? sql` AND session_id = ${options.sessionId}` : sql``}${options.type !== void 0 ? sql` AND type = ${options.type}` : sql``}${options.since !== void 0 ? sql` AND occurred_at >= ${options.since}` : sql``} ORDER BY occurred_at DESC, turn_no DESC LIMIT ${limit}`).map((r) => ({
+		const sessionFilter = options.sessionId !== void 0 ? sql` AND session_id = ${options.sessionId}` : sql``;
+		const typeFilter = options.type !== void 0 ? sql` AND type = ${options.type}` : sql``;
+		const sinceFilter = options.since !== void 0 ? sql` AND occurred_at >= ${options.since}` : sql``;
+		return (yield* sql`SELECT id, session_id, turn_no, type, payload, occurred_at FROM turns WHERE 1=1${sessionFilter}${typeFilter}${sinceFilter} ORDER BY occurred_at DESC, turn_no DESC LIMIT ${limit}`).map((r) => ({
 			id: r.id,
 			sessionId: r.session_id,
 			turnNo: r.turn_no,
@@ -1551,6 +1571,7 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 						WHERE project = ${options.project}
 						ORDER BY timestamp DESC LIMIT 1`;
 			if (runs.length === 0) return [];
+			const runId = runs[0].id;
 			return (yield* sql`
 						SELECT tc.id, tc.full_name, tc.state, tc.duration, f.path AS file_path, tc.classification
 						FROM test_cases tc
@@ -1558,7 +1579,7 @@ const DataReaderLive = Layer.effect(DataReader, Effect.gen(function* () {
 						JOIN files f ON f.id = tm.file_id
 						JOIN test_case_tags tct ON tct.test_case_id = tc.id
 						JOIN tags t ON t.id = tct.tag_id
-						WHERE tm.run_id = ${runs[0].id} AND t.name = ${tag}
+						WHERE tm.run_id = ${runId} AND t.name = ${tag}
 						ORDER BY tc.full_name
 					`).map(mapTestListRow);
 		}

package/layers/DataStoreLive.js

@@ -174,13 +174,14 @@ const DataStoreLive = Layer.effect(DataStore, Effect.gen(function* () {
 		table: "file_coverage",
 		reason: extractSqlReason(e)
 	})));
-	const writeHistory = (project, fullName, runId, timestamp, state, duration, flaky, retryCount, errorMessage) => Effect.gen(function* () {
+	const writeHistory = (project, fullName, modulePath, runId, timestamp, state, duration, flaky, retryCount, errorMessage) => Effect.gen(function* () {
 		yield* Effect.logDebug("writeHistory").pipe(Effect.annotateLogs({
 			project,
+			modulePath,
 			runId
 		}));
-		yield* sql`INSERT INTO test_history (run_id, project, full_name, timestamp, state, duration, flaky, retry_count, error_message) VALUES (${runId}, ${project}, ${fullName}, ${timestamp}, ${state}, ${duration}, ${flaky ? 1 : 0}, ${retryCount}, ${errorMessage})`;
-		yield* sql`DELETE FROM test_history WHERE id NOT IN (SELECT id FROM test_history WHERE project = ${project} AND full_name = ${fullName} ORDER BY timestamp DESC LIMIT 10) AND project = ${project} AND full_name = ${fullName}`;
+		yield* sql`INSERT INTO test_history (run_id, project, module_path, full_name, timestamp, state, duration, flaky, retry_count, error_message) VALUES (${runId}, ${project}, ${modulePath}, ${fullName}, ${timestamp}, ${state}, ${duration}, ${flaky ? 1 : 0}, ${retryCount}, ${errorMessage})`;
+		yield* sql`DELETE FROM test_history WHERE id NOT IN (SELECT id FROM test_history WHERE project = ${project} AND module_path = ${modulePath} AND full_name = ${fullName} ORDER BY timestamp DESC LIMIT 10) AND project = ${project} AND module_path = ${modulePath} AND full_name = ${fullName}`;
 	}).pipe(Effect.annotateLogs("service", "DataStore"), Effect.mapError((e) => new DataStoreError({
 		operation: "write",
 		table: "test_history",
@@ -226,7 +227,8 @@ const DataStoreLive = Layer.effect(DataStore, Effect.gen(function* () {
 			testModuleId,
 			mappingType
 		}));
-		yield* sql`INSERT OR IGNORE INTO source_test_map (source_file_id, test_module_id, mapping_type) VALUES (${yield* ensureFile(sourceFilePath)}, ${testModuleId}, ${mappingType})`;
+		const sourceFileId = yield* ensureFile(sourceFilePath);
+		yield* sql`INSERT OR IGNORE INTO source_test_map (source_file_id, test_module_id, mapping_type) VALUES (${sourceFileId}, ${testModuleId}, ${mappingType})`;
 	}).pipe(Effect.annotateLogs("service", "DataStore"), Effect.mapError((e) => new DataStoreError({
 		operation: "write",
 		table: "source_test_map",
@@ -342,11 +344,12 @@ const DataStoreLive = Layer.effect(DataStore, Effect.gen(function* () {
 			if (input.type === "file_edit") {
 				const filePath = payload.file_path;
 				yield* sql`INSERT OR IGNORE INTO files (path) VALUES (${filePath})`;
+				const fileId = (yield* sql`SELECT id FROM files WHERE path = ${filePath}`)[0].id;
 				yield* sql`
 									INSERT INTO file_edits (turn_id, file_id, edit_kind, lines_added, lines_removed, diff)
 									VALUES (
 										${turnId},
-										${(yield* sql`SELECT id FROM files WHERE path = ${filePath}`)[0].id},
+										${fileId},
 										${payload.edit_kind},
 										${payload.lines_added ?? null},
 										${payload.lines_removed ?? null},
@@ -355,11 +358,12 @@ const DataStoreLive = Layer.effect(DataStore, Effect.gen(function* () {
 								`;
 			} else {
 				const rawToolName = payload.tool_name;
+				const toolName = typeof rawToolName === "string" && rawToolName.startsWith("mcp__") ? rawToolName.split("__").at(-1) ?? rawToolName : rawToolName;
 				yield* sql`
 									INSERT INTO tool_invocations (turn_id, tool_name, params_hash, result_summary, duration_ms, success)
 									VALUES (
 										${turnId},
-										${typeof rawToolName === "string" && rawToolName.startsWith("mcp__") ? rawToolName.split("__").at(-1) ?? rawToolName : rawToolName},
+										${toolName},
 										${null},
 										${payload.result_summary ?? null},
 										${payload.duration_ms ?? null},
@@ -529,7 +533,7 @@ const DataStoreLive = Layer.effect(DataStore, Effect.gen(function* () {
 			goal: input.goal
 		}));
 		yield* ensureTddTaskOpen(input.tddTaskId);
-		return goalRowFromDb((yield* sql`
+		const rows = yield* sql`
 					INSERT INTO tdd_session_goals (session_id, ordinal, goal)
 					SELECT ${input.tddTaskId},
 					       COALESCE(MAX(ordinal), -1) + 1,
@@ -541,7 +545,8 @@ const DataStoreLive = Layer.effect(DataStore, Effect.gen(function* () {
 			operation: "write",
 			table: "tdd_session_goals",
 			reason: extractSqlReason(e)
-		}))))[0]);
+		})));
+		return goalRowFromDb(rows[0]);
 	}).pipe(Effect.annotateLogs("service", "DataStore"));
 	const getGoal = (id) => Effect.gen(function* () {
 		const rows = yield* sql`

package/layers/HistoryTrackerLive.js

@@ -1,5 +1,5 @@
 import { DataReader } from "../services/DataReader.js";
-import { HistoryTracker } from "../services/HistoryTracker.js";
+import { HistoryTracker, historyKey } from "../services/HistoryTracker.js";
 import { classifyTest } from "../utils/classify-test.js";
 import { Effect, Layer } from "effect";
 
@@ -11,26 +11,28 @@ const HistoryTrackerLive = Layer.effect(HistoryTracker, Effect.gen(function* ()
 	return { classify: (project, testOutcomes, timestamp) => Effect.gen(function* () {
 		const existing = yield* reader.getHistory(project);
 		const testMap = /* @__PURE__ */ new Map();
-		for (const entry of existing.tests) testMap.set(entry.fullName, {
+		for (const entry of existing.tests) testMap.set(historyKey(entry.modulePath, entry.fullName), {
 			...entry,
 			runs: [...entry.runs]
 		});
 		const classifications = /* @__PURE__ */ new Map();
 		for (const outcome of testOutcomes) {
-			let entry = testMap.get(outcome.fullName);
+			const key = historyKey(outcome.modulePath, outcome.fullName);
+			let entry = testMap.get(key);
 			if (!entry) {
 				entry = {
+					modulePath: outcome.modulePath,
 					fullName: outcome.fullName,
 					runs: []
 				};
-				testMap.set(outcome.fullName, entry);
+				testMap.set(key, entry);
 			}
 			const priorRuns = entry.runs;
 			entry.runs = [{
 				timestamp,
 				state: outcome.state
 			}, ...priorRuns].slice(0, WINDOW_SIZE);
-			classifications.set(outcome.fullName, classifyTest(outcome.state, priorRuns));
+			classifications.set(key, classifyTest(outcome.state, priorRuns));
 		}
 		return {
 			history: {

package/migrations/0001_initial.js

@@ -352,6 +352,7 @@ const migration = Effect.gen(function* () {
 			id INTEGER PRIMARY KEY AUTOINCREMENT,
 			run_id INTEGER NOT NULL REFERENCES test_runs(id) ON DELETE CASCADE,
 			project TEXT NOT NULL,
+			module_path TEXT NOT NULL,
 			full_name TEXT NOT NULL,
 			timestamp TEXT NOT NULL,
 			state TEXT NOT NULL CHECK (state IN ('passed', 'failed', 'skipped', 'pending')),
@@ -359,10 +360,10 @@ const migration = Effect.gen(function* () {
 			flaky INTEGER CHECK (flaky IN (0, 1)),
 			retry_count INTEGER DEFAULT 0,
 			error_message TEXT,
-			UNIQUE(project, full_name, timestamp)
+			UNIQUE(project, module_path, full_name, timestamp)
 		)
 	`;
-	yield* sql`CREATE INDEX idx_test_history_lookup ON test_history(project, full_name)`;
+	yield* sql`CREATE INDEX idx_test_history_lookup ON test_history(project, module_path, full_name)`;
 	yield* sql`CREATE INDEX idx_test_history_full_name ON test_history(full_name, timestamp)`;
 	yield* sql`CREATE INDEX idx_test_history_run ON test_history(run_id)`;
 	yield* sql`

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@vitest-agent/sdk",
-	"version": "1.2.0",
+	"version": "1.3.1",
 	"private": false,
 	"description": "Shared Effect Schema, SQLite migrations, data layer, services, formatters, and utilities for vitest-agent packages.",
 	"keywords": [
@@ -52,11 +52,11 @@
 		"acorn": "^8.17.0",
 		"acorn-typescript": "^1.4.13",
 		"ajv": "^8.20.0",
-		"config-file-effect": "^0.2.3",
+		"config-file-effect": "^0.3.0",
 		"effect": "^3.21.4",
 		"std-env": "^4.1.0",
-		"workspaces-effect": "^1.2.0",
-		"xdg-effect": "^2.0.1"
+		"workspaces-effect": "^2.0.2",
+		"xdg-effect": "^2.1.0"
 	},
 	"engines": {
 		"node": ">=24.11.0"

package/schemas/History.js

@@ -14,6 +14,7 @@ const TestRun = Schema.Struct({
 * @public
 */
 const TestHistory = Schema.Struct({
+	modulePath: Schema.String,
 	fullName: Schema.String,
 	runs: Schema.Array(TestRun)
 }).annotations({ identifier: "TestHistory" });

package/services/HistoryTracker.js

@@ -1,8 +1,20 @@
 import { Context } from "effect";
 
 //#region src/services/HistoryTracker.ts
+/**
+* Builds the composite (modulePath, fullName) key used to key the internal
+* testMap and the returned classifications Map, so identically-named tests
+* in different files classify independently instead of colliding.
+*
+* Uses `JSON.stringify` for an injective encoding: a plain delimiter such as
+* a space is not collision-proof (e.g. `("a b", "c")` and `("a", "b c")` would
+* both yield `"a b c"`), which would reintroduce the very collision class this
+* key exists to prevent.
+* @public
+*/
+const historyKey = (modulePath, fullName) => JSON.stringify([modulePath, fullName]);
 /** @public */
 var HistoryTracker = class extends Context.Tag("vitest-agent/HistoryTracker")() {};
 
 //#endregion
-export { HistoryTracker };
+export { HistoryTracker, historyKey };

package/services/idempotency.js

@@ -25,7 +25,8 @@ const IDEMPOTENCY_ROOT_SENTINEL = "__ROOT__";
 const deriveIdempotencyKey = (input) => {
 	const parent = input.parentAgentId ?? "__ROOT__";
 	const message = `${input.agentType}|${parent}|${input.clientNonce}`;
-	return base32Encode(createHash("sha256").update(message).digest()).slice(0, 26);
+	const digest = createHash("sha256").update(message).digest();
+	return base32Encode(digest).slice(0, 26);
 };
 const BASE32_ALPHABET = "abcdefghijklmnopqrstuvwxyz234567";
 /**

package/sql/assemblers.js

@@ -1,3 +1,5 @@
+import { historyKey } from "../services/HistoryTracker.js";
+
 //#region src/sql/assemblers.ts
 /**
 * Builds a `CacheManifest` from an array of test_runs aggregation rows.
@@ -32,10 +34,13 @@ function assembleManifest(rows, dbPath) {
 	};
 }
 /**
-* Groups test_history rows by `full_name` into a keyed record.
+* Groups test_history rows by the composite (module_path, full_name) key
+* into a keyed record.
 *
-* Returns `Record<fullName, { runs: TestRun[] }>` where `runs` are the
-* accumulated test runs for that test, in the order they appear in `rows`.
+* Returns `Record<compositeKey, { modulePath, fullName, runs: TestRun[] }>`
+* where `runs` are the accumulated test runs for that test, in the order
+* they appear in `rows`. Two rows sharing `full_name` but differing
+* `module_path` are tracked as distinct entries rather than merged.
 *
 * Only rows with state `"passed"` or `"failed"` are included in runs
 * (matching the `TestRun` schema constraint).
@@ -46,8 +51,13 @@ function assembleHistoryRecord(rows) {
 	for (const row of rows) {
 		const state = row.state === "passed" || row.state === "failed" ? row.state : null;
 		if (state === null) continue;
-		if (!record[row.full_name]) record[row.full_name] = { runs: [] };
-		record[row.full_name].runs.push({
+		const key = historyKey(row.module_path, row.full_name);
+		if (!record[key]) record[key] = {
+			modulePath: row.module_path,
+			fullName: row.full_name,
+			runs: []
+		};
+		record[key].runs.push({
 			timestamp: row.timestamp,
 			state
 		});

package/testing.d.ts

@@ -742,7 +742,7 @@ declare const DataStore_base: Context.TagClass<DataStore, "vitest-agent/DataStor
   readonly writeTestCases: (moduleId: number, tests: ReadonlyArray<TestCaseInput>) => Effect.Effect<ReadonlyArray<number>, DataStoreError>;
   readonly writeErrors: (runId: number, errors: ReadonlyArray<TestErrorInput>) => Effect.Effect<void, DataStoreError>;
   readonly writeCoverage: (runId: number, coverage: ReadonlyArray<FileCoverageInput>) => Effect.Effect<void, DataStoreError>;
-  readonly writeHistory: (project: string, fullName: string, runId: number, timestamp: string, state: string, duration: number | null, flaky: boolean, retryCount: number, errorMessage: string | null) => Effect.Effect<void, DataStoreError>;
+  readonly writeHistory: (project: string, fullName: string, modulePath: string, runId: number, timestamp: string, state: string, duration: number | null, flaky: boolean, retryCount: number, errorMessage: string | null) => Effect.Effect<void, DataStoreError>;
   readonly writeBaselines: (baselines: CoverageBaselines) => Effect.Effect<void, DataStoreError>;
   readonly writeTrends: (project: string, runId: number, entry: TrendEntry) => Effect.Effect<void, DataStoreError>;
   readonly writeSourceMap: (sourceFilePath: string, testModuleId: number, mappingType: string) => Effect.Effect<void, DataStoreError>;
@@ -1102,6 +1102,7 @@ declare const HistoryRecord: Schema.Struct<{
   project: typeof Schema.String;
   updatedAt: typeof Schema.String;
   tests: Schema.Array$<Schema.Struct<{
+    modulePath: typeof Schema.String;
     fullName: typeof Schema.String;
     runs: Schema.Array$<Schema.Struct<{
       timestamp: typeof Schema.String;
@@ -1126,6 +1127,7 @@ interface ProjectRunSummary {
 /** @public */
 interface FlakyTest {
   readonly fullName: string;
+  readonly modulePath: string;
   readonly project: string;
   readonly passCount: number;
   readonly failCount: number;
@@ -1135,6 +1137,7 @@ interface FlakyTest {
 /** @public */
 interface PersistentFailure {
   readonly fullName: string;
+  readonly modulePath: string;
   readonly project: string;
   readonly consecutiveFailures: number;
   readonly firstFailedAt: string;

package/testing.js

@@ -176,9 +176,9 @@ function flaky(filename) {
 			state: "passed",
 			duration: 300
 		}]);
-		yield* store.writeHistory("default", "async > resolves within timeout", runId1, "2025-12-31T23:00:00.000Z", "passed", 500, false, 0, null);
-		yield* store.writeHistory("default", "async > resolves within timeout", runId1, "2026-01-01T00:00:00.000Z", "failed", 500, false, 0, null);
-		yield* store.writeHistory("default", "async > rejects on error", runId1, "2026-01-01T00:00:00.000Z", "passed", 300, false, 0, null);
+		yield* store.writeHistory("default", "async > resolves within timeout", "src/async.test.ts", runId1, "2025-12-31T23:00:00.000Z", "passed", 500, false, 0, null);
+		yield* store.writeHistory("default", "async > resolves within timeout", "src/async.test.ts", runId1, "2026-01-01T00:00:00.000Z", "failed", 500, false, 0, null);
+		yield* store.writeHistory("default", "async > rejects on error", "src/async.test.ts", runId1, "2026-01-01T00:00:00.000Z", "passed", 300, false, 0, null);
 		const runId2 = yield* store.writeRun({
 			invocationId: "inv-preset-flaky-2",
 			project: "default",
@@ -211,8 +211,8 @@ function flaky(filename) {
 			state: "passed",
 			duration: 300
 		}]);
-		yield* store.writeHistory("default", "async > resolves within timeout", runId2, "2026-01-01T01:00:00.000Z", "passed", 400, false, 0, null);
-		yield* store.writeHistory("default", "async > rejects on error", runId2, "2026-01-01T01:00:00.000Z", "passed", 300, false, 0, null);
+		yield* store.writeHistory("default", "async > resolves within timeout", "src/async.test.ts", runId2, "2026-01-01T01:00:00.000Z", "passed", 400, false, 0, null);
+		yield* store.writeHistory("default", "async > rejects on error", "src/async.test.ts", runId2, "2026-01-01T01:00:00.000Z", "passed", 300, false, 0, null);
 	});
 	return Layer.effectDiscard(seed).pipe(Layer.provideMerge(base));
 }