oh-my-knowledge 1.0.0-beta.4 → 1.0.0-beta.5

package/README.md

@@ -22,6 +22,16 @@ Observe real-world performance, measure version differences, and determine wheth
 
 📖 **Full documentation: [oh-my-knowledge.pages.dev](https://oh-my-knowledge.pages.dev)** (searchable, English / 简体中文)
 
+## What knowledge means in OMK
+
+**Entities are the things knowledge describes; knowledge expresses their states, relationships, or actions in a specific context.**
+
+**How knowledge is expressed:**
+
+> **Time scope + scenario + conditions + entity A + relation or action + entity B (if any)**
+
+See [How OMK understands knowledge](docs/explanation/knowledge.md) for entity roles, shared relationships, and the distinction between knowledge and its carriers. Entity-based retrieval remains a design direction.
+
 ## What OMK helps you know
 
 | Decision | Command | Evidence you get |
@@ -35,6 +45,8 @@ Observe real-world performance, measure version differences, and determine wheth
 
 ## Quick start
 
+Already have a Node.js service, retriever, or Agent? Start with [Use OMK in your service](docs/guides/eval-runtime.md): run an example without model credentials, choose a scorer, connect your service, and read the results. The command-line quickstart follows below.
+
 ```bash
 npm i -g oh-my-knowledge
 omk init demo && cd demo
@@ -228,7 +240,7 @@ The full docs are published at **[oh-my-knowledge.pages.dev](https://oh-my-knowl
 - **[CLI reference](docs/reference/cli.md)** — all top-level commands with bash examples and flag tables
 - **[Migrate to the 1.0 preview](docs/guides/v1-preview-migration.md)** — install channel, storage reset, sample protocol, CLI automation, and embedded API changes since 0.54
 - **[Evaluation Core cutover](docs/guides/eval-core-cutover.md)** — `BREAKING-SCHEMA` storage, resume, Studio, Gold, managed-evidence, and evolve migration
-- **[Embed OMK in a service](docs/guides/eval-runtime.md)** — canonical `evaluate()` plus repeat-run `evaluateSeries()` for Node.js and FaaS hosts
+- **[Embed OMK in a service](docs/guides/eval-runtime.md)** — choose a scorer, connect a Node.js service, and interpret scores and failures
 - **[Storage layout v2](docs/specs/storage-layout-spec.md)** — project/global domains, compatibility boundary, and Git policy
 - **[Executors](docs/reference/executors.md)** & **[artifact layout](docs/reference/artifact-layout.md)** — built-in / custom executors; how `variant` resolves to an artifact + runtime context
 - **[How-to guides](docs/guides/agent-eval.md)** — [evaluate an agent](docs/guides/agent-eval.md) (project runtime context) and [use non-Claude models](docs/guides/non-claude-models.md) (GLM / Qwen / DeepSeek / Moonshot / Ollama)

package/README.zh.md

@@ -22,6 +22,16 @@
 
 📖 **完整文档：[oh-my-knowledge.pages.dev/zh](https://oh-my-knowledge.pages.dev/zh/)**（可搜索，可切换英文）
 
+## OMK 中的知识是什么
+
+**实体是知识所描述的事物；知识表达实体在特定上下文中的状态、关系或行为。**
+
+**知识的表达形式：**
+
+> **时间范围＋场景＋条件＋实体 A＋关系或行为＋实体 B（如有）**
+
+实体角色、多实体关联及知识与载体的区别，见[OMK 如何理解知识](docs/zh/explanation/knowledge.md)。实体关联检索仍是设计方向。
+
 ## OMK 让你知道什么
 
 | 决策问题 | 命令 | 你会得到的证据 |
@@ -35,6 +45,8 @@
 
 ## 快速开始
 
+已有 Node.js 服务、检索系统或 Agent，希望接入评分与版本对比？从[在服务中使用 OMK](docs/zh/guides/eval-runtime.md)开始：先运行无需模型凭证的示例，再选择评分方法、接入自己的服务并解读结果。下面是命令行上手流程。
+
 ```bash
 npm i -g oh-my-knowledge
 omk init demo && cd demo
@@ -228,7 +240,7 @@ omk-mcp
 - **[CLI 参考](docs/zh/reference/cli.md)** —— 顶层命令的 bash 示例和 flag 表
 - **[迁移到 1.0 预览版](docs/zh/guides/v1-preview-migration.md)** —— 从 `0.54` 升级时的安装渠道、存储重建、用例协议、CLI 自动化与嵌入式 API 变化
 - **[Evaluation Core 生产切换](docs/zh/guides/eval-core-cutover.md)** —— `BREAKING-SCHEMA` 存储、resume、Studio、Gold、受管证据与 evolve 迁移
-- **[在服务中嵌入 OMK](docs/zh/guides/eval-runtime.md)** —— 面向 Node.js 与 FaaS 宿主的 canonical `evaluate()` 与重复运行 `evaluateSeries()`
+- **[在服务中嵌入 OMK](docs/zh/guides/eval-runtime.md)** —— 选择评分方法、接入 Node.js 服务、解读分数与失败原因
 - **[存储布局 v2](docs/zh/specs/storage-layout-spec.md)** —— 项目／全局领域、迁移兼容与 Git 策略
 - **[执行器](docs/zh/reference/executors.md)** & **[知识载体布局](docs/zh/reference/artifact-layout.md)** —— 内置 / 自定义执行器；variant 如何解析为 artifact + runtime context
 - **[操作指南](docs/zh/guides/agent-eval.md)** —— [评测 agent](docs/zh/guides/agent-eval.md)（项目 runtime context）与[使用非 Claude 模型](docs/zh/guides/non-claude-models.md)（GLM / 通义 / DeepSeek / Moonshot / Ollama）

package/dist/cli/commands/eval/index.js

@@ -160,6 +160,10 @@ async function runEval(_args, flags, lang) {
             process.stdout.write(`Core: ${outcome.status?.runStatus ?? 'prepared'}／${outcome.status?.evidenceStatus ?? 'n/a'}／${outcome.status?.conclusionStatus ?? 'n/a'}\n`
                 + `Gate: ${outcome.gate?.gateStatus ?? 'not-applicable'}${outcome.gate?.reasonCodes?.length ? `（${outcome.gate.reasonCodes.join(', ')}）` : ''}\n`);
         }
+        // A blocked gate exits immediately through oclif; flush a piped JSON report first.
+        await new Promise((resolve, reject) => {
+            process.stdout.write('', (error) => error ? reject(error) : resolve());
+        });
         throw new CliExit(result.exitCode);
     }
     catch (err) {

package/dist/cli/lib/run-core-evaluation.js

@@ -2,7 +2,9 @@ import { existsSync } from 'node:fs';
 import { join, resolve } from 'node:path';
 import { compileCliEvaluationInput, parseCliEvaluationRequest, } from '../../eval-workflows/input-compilation/index.js';
 import { createNodeCoreBatchArtifactStore, createNodeCoreContentStore, createNodeCoreRunArtifactStore, createOverlayCoreRunArtifactStore, } from '../../eval-workflows/artifact-store/index.js';
-import { createNodeCliProductionComposition, createProductionEvaluationHost, executeProductionEvaluationSeries, persistCoreArtifactSidecars, resolveNodeCliEvaluationRequest, } from '../../eval-workflows/production-host/index.js';
+import { createNodeCliProductionComposition, } from '../../eval-hosts/node/node-cli-composition.js';
+import { createProductionEvaluationHost, executeProductionEvaluationSeries, persistCoreArtifactSidecars, } from '../../eval-workflows/production-host/index.js';
+import { resolveNodeCliEvaluationRequest, } from '../../eval-hosts/node/node-cli-evaluation-resolver.js';
 import { projectCoreCliDryRun, projectCoreCliBatchOutcome, projectCoreCliRunOutcome, projectCoreCliSeriesOutcome, } from '../../eval-workflows/projections/cli.js';
 import { projectCoreManagedEvidence } from '../../eval-workflows/projections/managed.js';
 import { discoverBatchSkills } from '../../eval-workflows/inputs/skill-loader.js';
@@ -226,10 +228,11 @@ export async function runCoreEvaluationCommand(input) {
         resourceLeaseRoot: machineLayout.resourceLeasesDir,
         environment: input.environment,
     });
-    const store = input.store ?? runStoreForOutput(outputDirectory, composition.support.contentResolver);
+    const store = input.store ?? runStoreForOutput(outputDirectory, composition.contentResolver);
     const host = {
         compiled,
-        ...composition,
+        runtime: composition.runtime,
+        schemaValidators: composition.schemaValidators,
         artifactStore: store,
     };
     if (compiled.orchestration.dryRun) {
@@ -261,8 +264,8 @@ export async function runCoreEvaluationCommand(input) {
         const memberArtifacts = await Promise.all(series.members.map(async (member) => {
             if (member.executionStatus !== 'started')
                 throw member.error;
-            await member.run.result;
             const persistence = await member.run.persistence;
+            await member.run.result;
             if (persistence.persistenceStatus !== 'stored') {
                 if (persistence.persistenceStatus === 'failed')
                     throw persistence.error;
@@ -316,8 +319,8 @@ export async function runCoreEvaluationCommand(input) {
             createdAt: new Date().toISOString(),
             progressSink: emitProgress(input.lang),
         });
-        await run.result;
         const persistence = await run.persistence;
+        await run.result;
         if (persistence.persistenceStatus !== 'stored') {
             if (persistence.persistenceStatus === 'failed')
                 throw persistence.error;

package/dist/dsh-plugin/core-adapter.d.ts

@@ -1,7 +1,7 @@
 import { type EvaluationDefinition } from '../eval-core/contracts/index.js';
 import { type ExecutionExecutor } from '../eval-core/execution/index.js';
-import type { OmkBindingResourceLeaseAccess } from '../eval-workflows/runtime-adapter/resource-leases/types.js';
-import type { RuntimeBindingOf } from '../eval-workflows/runtime-adapter/types.js';
+import type { OmkBindingResourceLeaseAccess } from '../eval-hosts/runtime-adapter/resource-leases/types.js';
+import type { RuntimeBindingOf } from '../eval-hosts/runtime-adapter/types.js';
 import type { DshAgentLike, DshHostContextLike } from './host-executor.js';
 export { DSH_HOST_CORE_ADAPTER_IMPLEMENTATION_VERSION, createDshHostCoreSchemaValidators, } from './core-protocol.js';
 export declare const DEFAULT_DSH_HOST_CORE_MAX_INPUT_BYTES: number;

package/dist/dsh-plugin/core-adapter.js

@@ -2,7 +2,7 @@ import { randomUUID } from 'node:crypto';
 import { JsonValueSchema, RuntimeIdentitySchema, canonicalizeJson, deepFreezeCanonicalJson, digestCanonicalJson, } from '../eval-core/contracts/index.js';
 import { ExecutionPortFailure, } from '../eval-core/execution/index.js';
 import { createDshHostRuntimeFingerprint } from '../executors/core/runtime-fingerprint.js';
-import { captureClaudeCliRunState, captureClaudeCliTarget, disposeClaudeCliTrial, openClaudeCliTrial, } from '../eval-workflows/runtime-adapter/adapters/claude/resources.js';
+import { captureClaudeCliRunState, captureClaudeCliTarget, disposeClaudeCliTrial, openClaudeCliTrial, } from '../eval-hosts/runtime-adapter/adapters/claude/resources.js';
 import { createSameProcessExecutorAdapter } from '../eval-runtime/adapters/same-process.js';
 import { DSH_HOST_CORE_ADAPTER_IMPLEMENTATION_VERSION, dshHostCoreExecutorCapabilities, parseDshHostCoreResult, } from './core-protocol.js';
 import { supportsDshTraceEventType } from './trace-adapter.js';
@@ -716,7 +716,7 @@ export async function createDshHostCoreExecutorAdapter(input) {
                 }
                 runState.acquireTrial();
                 try {
-                    return openClaudeCliTrial(trial, runState, host.maxInputBytes, RESOURCE_PROFILE);
+                    return await openClaudeCliTrial(trial, runState, host.maxInputBytes, RESOURCE_PROFILE);
                 }
                 catch (error) {
                     await runState.releaseTrial();
@@ -726,8 +726,8 @@ export async function createDshHostCoreExecutorAdapter(input) {
             execute({ runState, trialState, attempt, scope }) {
                 return executeDshHost(host, target, runState, trialState, attempt, scope.operationIsolationKey);
             },
-            disposeTrial({ runState }) {
-                return disposeClaudeCliTrial(runState, RESOURCE_PROFILE);
+            disposeTrial({ runState, trialState }) {
+                return disposeClaudeCliTrial(runState, trialState, RESOURCE_PROFILE);
             },
             disposeRun({ runState }) {
                 return runState.requestDispose();

package/dist/dsh-plugin/core-command.js

@@ -1,9 +1,14 @@
+import { createOmkRuntimeProvider, createOmkEvaluationSchemaValidators } from '../eval-hosts/runtime-adapter/composition.js';
 import { join, resolve } from 'node:path';
 import { schemaIdentityKey, } from '../eval-core/contracts/index.js';
 import { compileCliEvaluationInput, parseCliEvaluationRequest, } from '../eval-workflows/input-compilation/index.js';
 import { createNodeCoreContentStore, createNodeCoreRunArtifactStore, } from '../eval-workflows/artifact-store/index.js';
 import { globalLayout, projectLayout } from '../evidence/storage/layout.js';
-import { createNodeEvaluationRuntimeSupportPorts, createNodeHostPreflightDeclarations, createProductionEvaluationHost, createProductionRuntimeFactoryRegistry, createJudgeProviderRuntimeIdentity, executeProductionEvaluationSeries, persistCoreArtifactSidecars, resolveNodeCliEvaluationRequest, } from '../eval-workflows/production-host/index.js';
+import { createNodeEvaluationRuntimeSupportPorts, createProductionRuntimeFactoryRegistry, } from '../eval-hosts/node/runtime-registry.js';
+import { createNodeHostPreflightDeclarations, } from '../eval-hosts/node/node-cli-composition.js';
+import { createProductionEvaluationHost, executeProductionEvaluationSeries, persistCoreArtifactSidecars, } from '../eval-workflows/production-host/index.js';
+import { createJudgeProviderRuntimeIdentity, } from '../eval-hosts/node/judge-provider-identity.js';
+import { resolveNodeCliEvaluationRequest, } from '../eval-hosts/node/node-cli-evaluation-resolver.js';
 import { projectCoreCliRunOutcome, projectCoreCliSeriesOutcome, } from '../eval-workflows/projections/cli.js';
 import { projectCoreManagedEvidence } from '../eval-workflows/projections/managed.js';
 import { managedDir, recordCoreEvalEvidence } from '../knowledge-artifacts/governance/index.js';
@@ -229,9 +234,11 @@ export async function runDshCoreEvaluation(input) {
     });
     const host = {
         compiled,
-        factories: factories({ compiled, host: input.host, parentAgent: input.parentAgent, projectRoot }),
-        support,
-        resources: { leaseRoot: machineLayout.resourceLeasesDir },
+        runtime: createOmkRuntimeProvider({
+            compiled, factories: factories({ compiled, host: input.host, parentAgent: input.parentAgent, projectRoot }),
+            support, resources: { leaseRoot: machineLayout.resourceLeasesDir },
+        }),
+        schemaValidators: createOmkEvaluationSchemaValidators(support.schemaValidators),
         artifactStore,
     };
     const independentSeries = compiled.orchestration.independentSeries;
@@ -254,8 +261,8 @@ export async function runDshCoreEvaluation(input) {
         const artifacts = await Promise.all(series.members.map(async (member) => {
             if (member.executionStatus !== 'started')
                 throw member.error;
-            await member.run.result;
             const persisted = await member.run.persistence;
+            await member.run.result;
             if (persisted.persistenceStatus !== 'stored') {
                 throw persisted.persistenceStatus === 'failed'
                     ? persisted.error
@@ -285,8 +292,8 @@ export async function runDshCoreEvaluation(input) {
         createdAt: new Date().toISOString(),
         signal: input.signal,
     });
-    await run.result;
     const persisted = await run.persistence;
+    await run.result;
     if (persisted.persistenceStatus !== 'stored') {
         throw persisted.persistenceStatus === 'failed'
             ? persisted.error

package/dist/dsh-plugin/core-protocol.d.ts

@@ -1,6 +1,6 @@
 import { type CoreSchemaValidator, type ExecutorCapabilities, type JsonValue, type UsageRecord } from '../eval-core/contracts/index.js';
 import { type DshHostRunResult } from './protocol.js';
-export declare const DSH_HOST_CORE_ADAPTER_IMPLEMENTATION_VERSION: "1.1.0";
+export declare const DSH_HOST_CORE_ADAPTER_IMPLEMENTATION_VERSION: "2.0.0";
 export interface ParsedDshHostCoreResult {
     readonly output?: string;
     readonly trace: JsonValue;

package/dist/dsh-plugin/core-protocol.js

@@ -1,10 +1,10 @@
 import { z } from 'zod';
 import { EXECUTOR_CAPABILITIES_SCHEMA_VERSION, ExecutorCapabilitiesSchema, JsonValueSchema, UsageRecordSchema, deepFreezeCanonicalJson, digestCanonicalJson, } from '../eval-core/contracts/index.js';
 import { ExecutionPortFailure } from '../eval-core/execution/index.js';
-import { SOURCE_NEUTRAL_TRACE_WITHOUT_MOCKS_SCHEMA_DESCRIPTOR, SOURCE_NEUTRAL_TRACE_SCHEMA_VERSION, SourceNeutralTraceWithoutMocksSchema, } from '../eval-workflows/runtime-adapter/source-neutral-trace.js';
+import { SOURCE_NEUTRAL_TRACE_WITHOUT_MOCKS_SCHEMA_DESCRIPTOR, SOURCE_NEUTRAL_TRACE_SCHEMA_VERSION, SourceNeutralTraceWithoutMocksSchema, } from '../eval-runtime/traces/source-neutral.js';
 import { buildDshHostResult } from './protocol.js';
 import { supportsDshTraceEventType } from './trace-adapter.js';
-export const DSH_HOST_CORE_ADAPTER_IMPLEMENTATION_VERSION = '1.1.0';
+export const DSH_HOST_CORE_ADAPTER_IMPLEMENTATION_VERSION = '2.0.0';
 const SCHEMA_DESCRIPTORS = {
     input: { valueKind: 'json-value' },
     output: { valueKind: 'string' },

package/dist/eval-hosts/node/index.d.ts

@@ -0,0 +1,6 @@
+export * from './node-cli-composition.js';
+export * from './node-cli-evaluation-resolver.js';
+export * from './node-sample-content-resolver.js';
+export * from './safe-http-content-resolver.js';
+export * from './runtime-registry.js';
+export * from './judge-provider-identity.js';

package/dist/eval-hosts/node/index.js

@@ -0,0 +1,6 @@
+export * from './node-cli-composition.js';
+export * from './node-cli-evaluation-resolver.js';
+export * from './node-sample-content-resolver.js';
+export * from './safe-http-content-resolver.js';
+export * from './runtime-registry.js';
+export * from './judge-provider-identity.js';

package/dist/{eval-workflows/production-host → eval-hosts/node}/node-cli-composition.d.ts

@@ -1,13 +1,13 @@
-import type { CliEvaluationCompileResult } from '../input-compilation/index.js';
-import type { OmkRuntimePreflightDeclaration } from '../runtime-adapter/index.js';
+import type { EvaluationRuntimeProvider } from '../../eval-runtime/provider.js';
+import { type CoreSchemaValidator } from '../../eval-core/contracts/index.js';
+import type { CliEvaluationCompileResult } from '../../eval-workflows/input-compilation/index.js';
+import type { OmkRuntimePreflightDeclaration } from '../runtime-adapter/types.js';
 import type { ClassifiedEnvironmentEntry } from '../runtime-adapter/adapters/shared/classified-environment.js';
-import { createNodeEvaluationRuntimeSupportPorts, createProductionRuntimeFactoryRegistry } from './runtime-registry.js';
+import { createNodeEvaluationRuntimeSupportPorts } from './runtime-registry.js';
 export interface NodeCliProductionComposition {
-    readonly factories: ReturnType<typeof createProductionRuntimeFactoryRegistry>;
-    readonly support: ReturnType<typeof createNodeEvaluationRuntimeSupportPorts>;
-    readonly resources: {
-        readonly leaseRoot: string;
-    };
+    readonly runtime: EvaluationRuntimeProvider;
+    readonly schemaValidators: ReadonlyMap<string, CoreSchemaValidator>;
+    readonly contentResolver: ReturnType<typeof createNodeEvaluationRuntimeSupportPorts>['contentResolver'];
 }
 export interface CreateNodeCliProductionCompositionInput {
     readonly compiled: CliEvaluationCompileResult;

package/dist/{eval-workflows/production-host → eval-hosts/node}/node-cli-composition.js

@@ -1,3 +1,4 @@
+import { createOmkRuntimeProvider, createOmkEvaluationSchemaValidators } from '../runtime-adapter/composition.js';
 import { access, readFile, realpath } from 'node:fs/promises';
 import { constants } from 'node:fs';
 import { delimiter, isAbsolute, join } from 'node:path';
@@ -6,7 +7,13 @@ import { checkDependencies } from '../../executors/preflight/dependencies.js';
 import { createExecutor } from '../../executors/index.js';
 import { resolveExecutorRuntimeFingerprint } from '../../executors/core/runtime-fingerprint.js';
 import { OmkUserFacingPreflightFailure } from '../runtime-adapter/preflight.js';
-import { createAnthropicApiCoreSchemaValidators, createClaudeCliCoreSchemaValidators, createClaudeSdkCoreSchemaValidators, createCodexCliCoreSchemaValidators, createCodexSdkCoreSchemaValidators, createCustomCommandCoreSchemaValidators, createOpenAIApiCoreSchemaValidators, customCommandExecutorCapabilities, } from '../runtime-adapter/adapters/index.js';
+import { createAnthropicApiCoreSchemaValidators, } from '../runtime-adapter/adapters/anthropic/protocol.js';
+import { createClaudeCliCoreSchemaValidators, } from '../runtime-adapter/adapters/claude/cli-protocol.js';
+import { createClaudeSdkCoreSchemaValidators, } from '../runtime-adapter/adapters/claude/sdk-protocol.js';
+import { createCodexCliCoreSchemaValidators, } from '../runtime-adapter/adapters/codex/cli-protocol.js';
+import { createCodexSdkCoreSchemaValidators, } from '../runtime-adapter/adapters/codex/sdk-protocol.js';
+import { createCustomCommandCoreSchemaValidators, customCommandExecutorCapabilities, } from '../runtime-adapter/adapters/custom/command.js';
+import { createOpenAIApiCoreSchemaValidators, } from '../runtime-adapter/adapters/openai/protocol.js';
 import { createJudgeProviderRuntimeIdentity } from './judge-provider-identity.js';
 import { createNodeEvaluationRuntimeSupportPorts, createProductionRuntimeFactoryRegistry, } from './runtime-registry.js';
 const CREDENTIAL_ENVIRONMENT = new Set([
@@ -165,7 +172,7 @@ function dependencyDoctor(compiled, environment, projectRoot) {
                     return;
                 const { runDoctor } = await import('../../knowledge-artifacts/doctor/index.js');
                 const { renderDoctorReportText } = await import('../../knowledge-artifacts/doctor/renderer.js');
-                const { tEvalWorkflowMessage } = await import('../messages.js');
+                const { tEvalWorkflowMessage } = await import('../../eval-workflows/messages.js');
                 const language = compiled.presentation.language;
                 const report = await runDoctor({
                     artifacts,
@@ -452,14 +459,16 @@ export async function createNodeCliProductionComposition(input) {
         ...baseSupport,
         schemaValidators: productionSchemaValidators(),
     });
+    const factories = createProductionRuntimeFactoryRegistry({
+        executorsByImplementationId: new Map(configurations),
+        resolveJudgeInvocation: judgeResolver(environment),
+    });
     return Object.freeze({
-        factories: createProductionRuntimeFactoryRegistry({
-            executorsByImplementationId: new Map(configurations),
-            resolveJudgeInvocation: judgeResolver(environment),
-        }),
-        support,
-        resources: Object.freeze({
-            leaseRoot: input.resourceLeaseRoot,
+        runtime: createOmkRuntimeProvider({
+            compiled: input.compiled, factories, support,
+            resources: { leaseRoot: input.resourceLeaseRoot },
         }),
+        schemaValidators: createOmkEvaluationSchemaValidators(support.schemaValidators),
+        contentResolver: support.contentResolver,
     });
 }

package/dist/{eval-workflows/production-host → eval-hosts/node}/node-cli-evaluation-resolver.d.ts

@@ -1,5 +1,5 @@
-import { type CliEvaluationRequest, type ResolvedCliEvaluationInput } from '../input-compilation/index.js';
-import { type SampleContentResolverSession } from './sample-content-resolution.js';
+import { type CliEvaluationRequest, type ResolvedCliEvaluationInput } from '../../eval-workflows/input-compilation/index.js';
+import { type SampleContentResolverSession } from '../../eval-workflows/production-host/sample-content-resolution.js';
 export interface ResolveNodeCliEvaluationRequestOptions {
     /** Absolute semantic root for relative CLI／eval.yaml locators. */
     readonly projectRoot: string;

package/dist/{eval-workflows/production-host → eval-hosts/node}/node-cli-evaluation-resolver.js

@@ -2,15 +2,15 @@ import { createHash } from 'node:crypto';
 import { chmod, lstat, mkdir, readFile, writeFile } from 'node:fs/promises';
 import { extname, isAbsolute, join, resolve } from 'node:path';
 import { canonicalizeJson, deepFreezeCanonicalJson, JsonValueSchema, } from '../../eval-core/contracts/index.js';
-import { loadSamples } from '../inputs/load-samples.js';
-import { resolveArtifacts } from '../inputs/skill-loader.js';
+import { loadSamples } from '../../eval-workflows/inputs/load-samples.js';
+import { resolveArtifacts } from '../../eval-workflows/inputs/skill-loader.js';
 import { MOCK_INTERCEPTION_PLAN_MEDIA_TYPE } from '../../eval-runtime/mock-interception.js';
-import { RESOLVED_CLI_EVALUATION_INPUT_SCHEMA_VERSION, RESOLVED_HOST_RESOURCES_SCHEMA_VERSION, CliEvaluationInputError, } from '../input-compilation/index.js';
+import { RESOLVED_CLI_EVALUATION_INPUT_SCHEMA_VERSION, RESOLVED_HOST_RESOURCES_SCHEMA_VERSION, CliEvaluationInputError, } from '../../eval-workflows/input-compilation/index.js';
 import { digestNodeFileResource, digestNodeTreeResource, } from '../runtime-adapter/resource-leases/node.js';
-import { BOOTSTRAP_FAMILY_ANALYSIS_V2_IMPLEMENTATION_ID, } from '../runtime-adapter/analysis/index.js';
-import { buildProductionMeasurementDesign } from './measurement-design.js';
+import { BOOTSTRAP_FAMILY_ANALYSIS_V2_IMPLEMENTATION_ID, } from '../../eval-workflows/measurement/analysis/bootstrap-family-node-v2.js';
+import { buildProductionMeasurementDesign } from '../../eval-workflows/production-host/measurement-design.js';
 import { createNodeSampleContentResolver } from './node-sample-content-resolver.js';
-import { SampleContentResolutionError, hasResolvableSampleUrls, resolveSampleContents, } from './sample-content-resolution.js';
+import { SampleContentResolutionError, hasResolvableSampleUrls, resolveSampleContents, } from '../../eval-workflows/production-host/sample-content-resolution.js';
 function fail(input) {
     throw new CliEvaluationInputError(input);
 }

package/dist/{eval-workflows/production-host → eval-hosts/node}/node-sample-content-resolver.d.ts

@@ -1,4 +1,4 @@
-import { type SampleContentResolution, type SampleContentResolverSession } from './sample-content-resolution.js';
+import { type SampleContentResolution, type SampleContentResolverSession } from '../../eval-workflows/production-host/sample-content-resolution.js';
 interface SampleContentMcpFetchTool {
     readonly name: string;
     readonly urlParam?: string;

package/dist/{eval-workflows/production-host → eval-hosts/node}/node-sample-content-resolver.js

@@ -1,7 +1,7 @@
 import { readFile } from 'node:fs/promises';
 import { Client } from '@modelcontextprotocol/sdk/client/index.js';
 import { StdioClientTransport } from '@modelcontextprotocol/sdk/client/stdio.js';
-import { SampleContentResolutionError, safeUrlLabel, } from './sample-content-resolution.js';
+import { SampleContentResolutionError, safeUrlLabel, } from '../../eval-workflows/production-host/sample-content-resolution.js';
 import { resolveSafeHttpSampleContent } from './safe-http-content-resolver.js';
 const MCP_TIMEOUT_MS = 30_000;
 function isRecord(value) {

package/dist/{eval-workflows/production-host → eval-hosts/node}/runtime-registry.d.ts

@@ -1,7 +1,13 @@
 import type { EvaluationEngineClock } from '../../eval-core/engine/index.js';
-import { type AnthropicApiCoreConfiguration, type ClaudeCliCoreConfiguration, type ClaudeSdkCoreConfiguration, type CodexCliCoreConfiguration, type CodexSdkCoreConfiguration, type CustomCommandConfiguration, type CustomCommandRuntimeDescription, type OpenAIApiCoreConfiguration } from '../runtime-adapter/adapters/index.js';
+import { type AnthropicApiCoreConfiguration } from '../runtime-adapter/adapters/anthropic/api.js';
+import { type ClaudeCliCoreConfiguration } from '../runtime-adapter/adapters/claude/cli.js';
+import { type ClaudeSdkCoreConfiguration } from '../runtime-adapter/adapters/claude/sdk.js';
+import { type CodexCliCoreConfiguration } from '../runtime-adapter/adapters/codex/cli.js';
+import { type CodexSdkCoreConfiguration } from '../runtime-adapter/adapters/codex/sdk.js';
+import { type CustomCommandConfiguration, type CustomCommandRuntimeDescription } from '../runtime-adapter/adapters/custom/command.js';
+import { type OpenAIApiCoreConfiguration } from '../runtime-adapter/adapters/openai/api.js';
 import type { OmkEvaluationRuntimeSupportPorts } from '../runtime-adapter/composition.js';
-import { type OmkLlmJudgeInvocationResolver } from '../runtime-adapter/evaluators/llm-assertions.js';
+import { type OmkLlmJudgeInvocationResolver } from '../runtime-adapter/evaluators/llm-assertion-factory.js';
 import type { OmkRuntimeBindingFactories, OmkRuntimePreflightDeclaration } from '../runtime-adapter/types.js';
 type ProductionExecutorPreflightConfiguration = {
     /** Physical checks are host-owned and captured without executing them. */

package/dist/{eval-workflows/production-host → eval-hosts/node}/runtime-registry.js

@@ -1,10 +1,17 @@
+import { LLM_ASSERTION_EVALUATOR_IMPLEMENTATION_ID } from '../../eval-workflows/measurement/evaluators/llm-assertions.js';
 import { isAbsolute } from 'node:path';
 import { createNodeEvaluationClock } from '../../eval-runtime/clock.js';
 import { deepFreezeCanonicalJson, } from '../../eval-core/contracts/index.js';
-import { createNodeCoreContentStore } from '../artifact-store/node-content-store.js';
-import { createAnthropicApiExecutorAdapter, createClaudeCliExecutorAdapter, createClaudeSdkExecutorAdapter, createCodexCliExecutorAdapter, createCodexSdkExecutorAdapter, createCustomCommandExecutorAdapter, createOpenAIApiExecutorAdapter, } from '../runtime-adapter/adapters/index.js';
+import { createNodeCoreContentStore } from '../../eval-workflows/artifact-store/node-content-store.js';
+import { createAnthropicApiExecutorAdapter, } from '../runtime-adapter/adapters/anthropic/api.js';
+import { createClaudeCliExecutorAdapter, } from '../runtime-adapter/adapters/claude/cli.js';
+import { createClaudeSdkExecutorAdapter, } from '../runtime-adapter/adapters/claude/sdk.js';
+import { createCodexCliExecutorAdapter, } from '../runtime-adapter/adapters/codex/cli.js';
+import { createCodexSdkExecutorAdapter, } from '../runtime-adapter/adapters/codex/sdk.js';
+import { createCustomCommandExecutorAdapter, } from '../runtime-adapter/adapters/custom/command.js';
+import { createOpenAIApiExecutorAdapter, } from '../runtime-adapter/adapters/openai/api.js';
 import { captureCoreApiTransport } from '../runtime-adapter/adapters/shared/api-http.js';
-import { createLlmAssertionEvaluatorBindingFactory, LLM_ASSERTION_EVALUATOR_IMPLEMENTATION_ID, } from '../runtime-adapter/evaluators/llm-assertions.js';
+import { createLlmAssertionEvaluatorBindingFactory, } from '../runtime-adapter/evaluators/llm-assertion-factory.js';
 import { createRubricJudgeEvaluatorBindingFactory, RUBRIC_JUDGE_EVALUATOR_IMPLEMENTATION_ID, } from '../runtime-adapter/evaluators/rubric-judge.js';
 export class ProductionRuntimeRegistryError extends TypeError {
     code;

package/dist/{eval-workflows/production-host → eval-hosts/node}/safe-http-content-resolver.d.ts

@@ -1,4 +1,4 @@
-import type { SampleContentResolution } from './sample-content-resolution.js';
+import type { SampleContentResolution } from '../../eval-workflows/production-host/sample-content-resolution.js';
 export declare function isPublicNetworkAddress(address: string, family: 4 | 6): boolean;
 /** Fetches public text while pinning validated DNS results across each request hop. */
 export declare function resolveSafeHttpSampleContent(urlString: string): Promise<SampleContentResolution>;

package/dist/{eval-workflows → eval-hosts}/runtime-adapter/adapters/anthropic/api.js

@@ -5,7 +5,7 @@ import { captureClassifiedEnvironment, mergeOutputClassification, } from '../sha
 import { ApiResponseBodyError, ApiResponseLimitError, captureCoreApiTransport, discardApiResponse, normalizeCoreApiEndpoint, readBoundedJsonResponse, requiredApiHeaderValue, } from '../shared/api-http.js';
 import { ANTHROPIC_API_CORE_ADAPTER_IMPLEMENTATION_VERSION, anthropicApiExecutorCapabilities, parseAnthropicApiMessage, } from './protocol.js';
 import { captureStatelessApiRunState, captureStatelessApiTarget, openStatelessApiTrial, } from '../shared/stateless-api-resources.js';
-import { createSameProcessExecutorAdapter } from '../shared/omk-resource-same-process.js';
+import { createSameProcessExecutorAdapter } from '../../../../eval-runtime/adapters/same-process.js';
 export { ANTHROPIC_API_CORE_ADAPTER_IMPLEMENTATION_VERSION, createAnthropicApiCoreSchemaValidators, } from './protocol.js';
 export const DEFAULT_ANTHROPIC_API_ENDPOINT = 'https://api.anthropic.com/v1/messages';
 export const DEFAULT_ANTHROPIC_API_MAX_REQUEST_BYTES = 2 * 1024 * 1024;

package/dist/{eval-workflows → eval-hosts}/runtime-adapter/adapters/anthropic/protocol.js

@@ -1,7 +1,7 @@
 import { UsageRecordSchema, } from '../../../../eval-core/contracts/index.js';
 import { ExecutionPortFailure } from '../../../../eval-core/execution/index.js';
 import { createStatelessApiCoreSchemaValidators, statelessApiExecutorCapabilities, } from '../shared/api-protocol-core.js';
-import { SOURCE_NEUTRAL_TRACE_SCHEMA_VERSION } from '../../source-neutral-trace.js';
+import { SOURCE_NEUTRAL_TRACE_SCHEMA_VERSION } from '../../../../eval-runtime/traces/source-neutral.js';
 export const ANTHROPIC_API_CORE_ADAPTER_IMPLEMENTATION_VERSION = '1.1.0';
 export const ANTHROPIC_API_PROTOCOL_PROFILE = Object.freeze({
     providerId: 'anthropic-api',

package/dist/{eval-workflows → eval-hosts}/runtime-adapter/adapters/claude/cli-protocol.d.ts

@@ -1,6 +1,6 @@
 import { type CoreSchemaValidator, type ExecutorCapabilities, type JsonValue, type UsageRecord } from '../../../../eval-core/contracts/index.js';
 import type { ClaudeMessage } from '../../../../executors/anthropic/claude/protocol.js';
-export declare const CLAUDE_CLI_CORE_ADAPTER_IMPLEMENTATION_VERSION: "1.2.0";
+export declare const CLAUDE_CLI_CORE_ADAPTER_IMPLEMENTATION_VERSION: "2.0.0";
 export interface ParsedClaudeCliStream {
     readonly messages: readonly ClaudeMessage[];
     readonly output?: string;

package/dist/{eval-workflows → eval-hosts}/runtime-adapter/adapters/claude/cli-protocol.js

@@ -2,8 +2,8 @@ import { z } from 'zod';
 import { EXECUTOR_CAPABILITIES_SCHEMA_VERSION, ExecutorCapabilitiesSchema, JsonValueSchema, UsageRecordSchema, deepFreezeCanonicalJson, digestCanonicalJson, } from '../../../../eval-core/contracts/index.js';
 import { ExecutionPortFailure } from '../../../../eval-core/execution/index.js';
 import { extractClaudeTrace, isClaudeResultMessage, } from '../../../../executors/anthropic/claude/trace.js';
-import { SOURCE_NEUTRAL_TRACE_SCHEMA_DESCRIPTOR, SOURCE_NEUTRAL_TRACE_SCHEMA_VERSION, SourceNeutralTraceSchema, } from '../../source-neutral-trace.js';
-export const CLAUDE_CLI_CORE_ADAPTER_IMPLEMENTATION_VERSION = '1.2.0';
+import { SOURCE_NEUTRAL_TRACE_SCHEMA_DESCRIPTOR, SOURCE_NEUTRAL_TRACE_SCHEMA_VERSION, SourceNeutralTraceSchema, } from '../../../../eval-runtime/traces/source-neutral.js';
+export const CLAUDE_CLI_CORE_ADAPTER_IMPLEMENTATION_VERSION = '2.0.0';
 const CLAUDE_SCHEMA_DESCRIPTORS = {
     input: { valueKind: 'json-value' },
     output: { valueKind: 'string' },

package/dist/{eval-workflows → eval-hosts}/runtime-adapter/adapters/claude/cli.js

@@ -9,8 +9,8 @@ import { captureClassifiedEnvironment, mergeOutputClassification, } from '../sha
 import { assertIdentityFilesUnchanged, captureIdentityFiles, } from '../shared/content-identity.js';
 import { CLAUDE_CLI_CORE_ADAPTER_IMPLEMENTATION_VERSION, claudeCliExecutorCapabilities, parseClaudeCliStream, } from './cli-protocol.js';
 import { captureClaudeCliRunState, captureClaudeCliTarget, disposeClaudeCliMockHandle, disposeClaudeCliTrial, openClaudeCliTrial, } from './resources.js';
-import { createSameProcessExecutorAdapter } from '../shared/omk-resource-same-process.js';
-import { attachSourceNeutralMockStats } from '../../source-neutral-trace.js';
+import { createSameProcessExecutorAdapter } from '../../../../eval-runtime/adapters/same-process.js';
+import { attachSourceNeutralMockStats } from '../../../../eval-runtime/traces/source-neutral.js';
 export { CLAUDE_CLI_CORE_ADAPTER_IMPLEMENTATION_VERSION, createClaudeCliCoreSchemaValidators, } from './cli-protocol.js';
 export const DEFAULT_CLAUDE_CLI_MAX_OUTPUT_BYTES = 10 * 1024 * 1024;
 export const DEFAULT_CLAUDE_CLI_MAX_INPUT_BYTES = 2 * 1024 * 1024;
@@ -585,7 +585,7 @@ export async function createClaudeCliExecutorAdapter(input) {
                 }
                 runState.acquireTrial();
                 try {
-                    return openClaudeCliTrial(trial, runState, configuration.maxInputBytes);
+                    return await openClaudeCliTrial(trial, runState, configuration.maxInputBytes);
                 }
                 catch (error) {
                     await runState.releaseTrial();
@@ -596,8 +596,8 @@ export async function createClaudeCliExecutorAdapter(input) {
                 await assertIdentityUnchanged(files, attempt.signal);
                 return executeAttempt(configuration, target, runState, trialState, attempt);
             },
-            disposeTrial({ runState }) {
-                return disposeClaudeCliTrial(runState);
+            disposeTrial({ runState, trialState }) {
+                return disposeClaudeCliTrial(runState, trialState);
             },
             disposeRun({ runState }) {
                 return runState.requestDispose();

package/dist/{eval-workflows → eval-hosts}/runtime-adapter/adapters/claude/resources.d.ts

@@ -85,6 +85,7 @@ export interface ClaudeCliRunState {
     requestDispose(): Promise<void>;
 }
 export interface ClaudeCliTrialState {
+    closeWorkspace(): Promise<void>;
     readonly prompt: string;
     readonly workingDirectory: string;
     readonly allowedTools?: readonly string[];
@@ -118,7 +119,7 @@ export declare const CLAUDE_SDK_RESOURCE_PROFILE: Readonly<{
 }>;
 export declare function captureClaudeCliTarget(targetInput: EvaluationDefinition['targets'][number], bindingInput: RuntimeBindingOf<'executor'>, profile?: ClaudeResourceProjectionProfile): CapturedClaudeCliTarget;
 export declare function captureClaudeCliRunState(lease: OmkBindingResourceLease, target: CapturedClaudeCliTarget, maxInputBytes: number, profile?: ClaudeResourceProjectionProfile): Promise<ClaudeCliRunState>;
-export declare function openClaudeCliTrial(trial: Readonly<ExecutorTrialContext>, runState: ClaudeCliRunState, maxInputBytes: number, profile?: ClaudeResourceProjectionProfile): ClaudeCliTrialState;
-export declare function disposeClaudeCliTrial(runState: ClaudeCliRunState, profile?: ClaudeResourceProjectionProfile): Promise<void>;
+export declare function openClaudeCliTrial(trial: Readonly<ExecutorTrialContext>, runState: ClaudeCliRunState, maxInputBytes: number, profile?: ClaudeResourceProjectionProfile): Promise<ClaudeCliTrialState>;
+export declare function disposeClaudeCliTrial(runState: ClaudeCliRunState, trialState: ClaudeCliTrialState, profile?: ClaudeResourceProjectionProfile): Promise<void>;
 export declare function disposeClaudeCliMockHandle(handle: CliMockHandle, profile?: ClaudeResourceProjectionProfile): Promise<void>;
 export {};

package/dist/{eval-workflows → eval-hosts}/runtime-adapter/adapters/claude/resources.js

@@ -1,3 +1,4 @@
+import { openNodeTrialWorkspace } from '../shared/trial-workspace.js';
 import { lstat, mkdtemp, readFile, rm } from 'node:fs/promises';
 import { tmpdir } from 'node:os';
 import { join, relative } from 'node:path';
@@ -576,7 +577,7 @@ export async function captureClaudeCliRunState(lease, target, maxInputBytes, pro
                 || workspace.leaseMode !== 'copy-on-write-overlay'
                 || !sameDescriptor(workspace, workspaceDescriptor))
                 fail(profile, 'WORKSPACE_INVALID', 'workspace lease does not match the sealed Target.');
-            workspaceDirectoriesByResourceId.set(workspaceDescriptor.resourceId, workspace.overlayPath);
+            workspaceDirectoriesByResourceId.set(workspaceDescriptor.resourceId, workspace.baseSnapshotPath);
         }
         privateWorkingDirectory = await mkdtemp(join(tmpdir(), 'omk-claude-run-'));
         let systemPromptBytes = 0;
@@ -649,13 +650,13 @@ export async function captureClaudeCliRunState(lease, target, maxInputBytes, pro
         fail(profile, 'RUN_MATERIALIZATION_FAILED', 'run state could not be materialized.');
     }
 }
-export function openClaudeCliTrial(trial, runState, maxInputBytes, profile = CLAUDE_CLI_RESOURCE_PROFILE) {
+export async function openClaudeCliTrial(trial, runState, maxInputBytes, profile = CLAUDE_CLI_RESOURCE_PROFILE) {
     if (canonicalizeJson(trial.executionControl) !== canonicalizeJson(resolveEffectiveExecutionControl(runState.executionControls, trial.sampleId))) {
         fail(profile, 'EXECUTION_CONTROL_MISMATCH', 'Trial control differs from the sealed Target.');
     }
     const workspace = trial.executionControl.workspace;
-    const workingDirectory = workspace.workspaceMode === 'not-required'
-        ? runState.privateWorkingDirectory
+    const baseSnapshotPath = workspace.workspaceMode === 'not-required'
+        ? undefined
         : runState.workspaceDirectoriesByResourceId.get(workspace.descriptor.resourceId)
             ?? fail(profile, 'WORKSPACE_INVALID', 'Trial workspace is absent from the sealed lease.');
     const toolControl = trial.executionControl.tools;
@@ -691,18 +692,29 @@ export function openClaudeCliTrial(trial, runState, maxInputBytes, profile = CLA
     if (Buffer.byteLength(prompt) + runState.systemPromptBytes > maxInputBytes) {
         fail(profile, 'INPUT_LIMIT_EXCEEDED', 'prompt exceeds the adapter input limit.');
     }
+    const trialWorkspace = await openNodeTrialWorkspace({
+        parentRoot: runState.privateWorkingDirectory,
+        ...(baseSnapshotPath === undefined ? {} : { baseSnapshotPath }),
+        signal: trial.signal,
+    });
     return Object.freeze({
         prompt,
-        workingDirectory,
+        workingDirectory: trialWorkspace.workingDirectory,
+        closeWorkspace: trialWorkspace.close,
         ...(allowedTools === undefined ? {} : { allowedTools: Object.freeze(allowedTools) }),
         ...(mockControls === undefined ? {} : { mocks: mockControls.mocks }),
         mocksStrict: mockControls?.strict ?? false,
         classification: higherClassification(runState.classification, mockControls?.classification ?? 'public'),
     });
 }
-export async function disposeClaudeCliTrial(runState, profile = CLAUDE_CLI_RESOURCE_PROFILE) {
+export async function disposeClaudeCliTrial(runState, trialState, profile = CLAUDE_CLI_RESOURCE_PROFILE) {
     try {
-        await runState.releaseTrial();
+        try {
+            await trialState.closeWorkspace();
+        }
+        finally {
+            await runState.releaseTrial();
+        }
     }
     catch {
         fail(profile, 'TRIAL_DISPOSE_FAILED', 'trial state could not be released.');

package/dist/{eval-workflows → eval-hosts}/runtime-adapter/adapters/claude/sdk-protocol.d.ts

@@ -1,6 +1,6 @@
 import { type CoreSchemaValidator, type ExecutorCapabilities } from '../../../../eval-core/contracts/index.js';
 import { type ParsedClaudeCliStream } from './cli-protocol.js';
-export declare const CLAUDE_SDK_CORE_ADAPTER_IMPLEMENTATION_VERSION: "1.2.0";
+export declare const CLAUDE_SDK_CORE_ADAPTER_IMPLEMENTATION_VERSION: "2.0.0";
 export type ParsedClaudeSdkStream = ParsedClaudeCliStream;
 export declare function createClaudeSdkCoreSchemaValidators(): readonly CoreSchemaValidator[];
 export declare function claudeSdkExecutorCapabilities(): ExecutorCapabilities;