@prismatic-io/lux 0.0.2-preview.3 → 0.0.2-preview.30
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/lib/answerers/claude-code/index.d.ts +15 -3
- package/lib/answerers/claude-code/index.d.ts.map +1 -1
- package/lib/answerers/claude-code/index.js +6 -5
- package/lib/answerers/claude-code/index.js.map +1 -1
- package/lib/answerers/persona/index.d.ts +44 -5
- package/lib/answerers/persona/index.d.ts.map +1 -1
- package/lib/answerers/persona/index.js +60 -34
- package/lib/answerers/persona/index.js.map +1 -1
- package/lib/answerers/terminal/index.d.ts +9 -3
- package/lib/answerers/terminal/index.d.ts.map +1 -1
- package/lib/answerers/terminal/index.js +7 -5
- package/lib/answerers/terminal/index.js.map +1 -1
- package/lib/assertions/authoring.d.ts +2 -2
- package/lib/assertions/authoring.d.ts.map +1 -1
- package/lib/assertions/authoring.js.map +1 -1
- package/lib/assertions/core/command-exits-zero.d.ts +3 -1
- package/lib/assertions/core/command-exits-zero.d.ts.map +1 -1
- package/lib/assertions/core/command-exits-zero.js +50 -7
- package/lib/assertions/core/command-exits-zero.js.map +1 -1
- package/lib/assertions/core/event-checks.d.ts +3 -0
- package/lib/assertions/core/event-checks.d.ts.map +1 -1
- package/lib/assertions/core/event-checks.js +49 -3
- package/lib/assertions/core/event-checks.js.map +1 -1
- package/lib/assertions/core/helpers.d.ts +2 -7
- package/lib/assertions/core/helpers.d.ts.map +1 -1
- package/lib/assertions/core/helpers.js +2 -6
- package/lib/assertions/core/helpers.js.map +1 -1
- package/lib/assertions/core/index.d.ts.map +1 -1
- package/lib/assertions/core/index.js +2 -1
- package/lib/assertions/core/index.js.map +1 -1
- package/lib/assertions/core/resource-checks.d.ts +19 -0
- package/lib/assertions/core/resource-checks.d.ts.map +1 -1
- package/lib/assertions/core/resource-checks.js +35 -5
- package/lib/assertions/core/resource-checks.js.map +1 -1
- package/lib/assertions/rubric/index.d.ts +26 -3
- package/lib/assertions/rubric/index.d.ts.map +1 -1
- package/lib/assertions/rubric/index.js +59 -13
- package/lib/assertions/rubric/index.js.map +1 -1
- package/lib/assertions/rubric/internal.d.ts +30 -5
- package/lib/assertions/rubric/internal.d.ts.map +1 -1
- package/lib/assertions/rubric/internal.js +181 -28
- package/lib/assertions/rubric/internal.js.map +1 -1
- package/lib/authoring.d.ts +36 -2
- package/lib/authoring.d.ts.map +1 -1
- package/lib/authoring.js +36 -1
- package/lib/authoring.js.map +1 -1
- package/lib/cli/bin.js +2 -13
- package/lib/cli/bin.js.map +1 -1
- package/lib/cli/command-runtime.d.ts +3 -3
- package/lib/cli/command-runtime.d.ts.map +1 -1
- package/lib/cli/command-runtime.js +17 -6
- package/lib/cli/command-runtime.js.map +1 -1
- package/lib/cli/doctor.d.ts +55 -0
- package/lib/cli/doctor.d.ts.map +1 -0
- package/lib/cli/doctor.js +14 -0
- package/lib/cli/doctor.js.map +1 -0
- package/lib/cli/harness-plan.d.ts +26 -0
- package/lib/cli/harness-plan.d.ts.map +1 -0
- package/lib/cli/harness-plan.js +75 -0
- package/lib/cli/harness-plan.js.map +1 -0
- package/lib/cli/init-templates.d.ts +11 -0
- package/lib/cli/init-templates.d.ts.map +1 -0
- package/lib/cli/init-templates.js +215 -0
- package/lib/cli/init-templates.js.map +1 -0
- package/lib/cli/init.d.ts +10 -0
- package/lib/cli/init.d.ts.map +1 -1
- package/lib/cli/init.js +34 -173
- package/lib/cli/init.js.map +1 -1
- package/lib/cli/output-schemas.d.ts +1406 -0
- package/lib/cli/output-schemas.d.ts.map +1 -0
- package/lib/cli/output-schemas.js +91 -0
- package/lib/cli/output-schemas.js.map +1 -0
- package/lib/cli/program.d.ts +140 -12
- package/lib/cli/program.d.ts.map +1 -1
- package/lib/cli/program.js +778 -646
- package/lib/cli/program.js.map +1 -1
- package/lib/cli/render/reporter.d.ts.map +1 -1
- package/lib/cli/render/reporter.js +40 -7
- package/lib/cli/render/reporter.js.map +1 -1
- package/lib/cli/run-options.d.ts +20 -4
- package/lib/cli/run-options.d.ts.map +1 -1
- package/lib/cli/run-options.js +100 -19
- package/lib/cli/run-options.js.map +1 -1
- package/lib/cli/view.d.ts +15 -0
- package/lib/cli/view.d.ts.map +1 -0
- package/lib/cli/view.js +168 -0
- package/lib/cli/view.js.map +1 -0
- package/lib/container/driver.d.ts +27 -0
- package/lib/container/driver.d.ts.map +1 -0
- package/lib/container/driver.js +197 -0
- package/lib/container/driver.js.map +1 -0
- package/lib/container/protocol.d.ts +12 -0
- package/lib/container/protocol.d.ts.map +1 -0
- package/lib/container/protocol.js +38 -0
- package/lib/container/protocol.js.map +1 -0
- package/lib/container/resources.d.ts +12 -0
- package/lib/container/resources.d.ts.map +1 -0
- package/lib/container/resources.js +133 -0
- package/lib/container/resources.js.map +1 -0
- package/lib/container/worker.d.ts +2 -0
- package/lib/container/worker.d.ts.map +1 -0
- package/lib/container/worker.js +85 -0
- package/lib/container/worker.js.map +1 -0
- package/lib/core/action-events.d.ts +29 -0
- package/lib/core/action-events.d.ts.map +1 -0
- package/lib/core/action-events.js +51 -0
- package/lib/core/action-events.js.map +1 -0
- package/lib/core/agent-activity.d.ts +49 -0
- package/lib/core/agent-activity.d.ts.map +1 -0
- package/lib/core/agent-activity.js +206 -0
- package/lib/core/agent-activity.js.map +1 -0
- package/lib/core/agent-environment.d.ts +5 -0
- package/lib/core/agent-environment.d.ts.map +1 -0
- package/lib/core/agent-environment.js +47 -0
- package/lib/core/agent-environment.js.map +1 -0
- package/lib/core/agent-events.d.ts +26 -0
- package/lib/core/agent-events.d.ts.map +1 -0
- package/lib/core/agent-events.js +13 -0
- package/lib/core/agent-events.js.map +1 -0
- package/lib/core/annotation.d.ts +6 -2
- package/lib/core/annotation.d.ts.map +1 -1
- package/lib/core/annotation.js +2 -1
- package/lib/core/annotation.js.map +1 -1
- package/lib/core/answerer.d.ts +12 -0
- package/lib/core/answerer.d.ts.map +1 -1
- package/lib/core/answerer.js +2 -0
- package/lib/core/answerer.js.map +1 -1
- package/lib/core/artifact-evidence.d.ts +8 -0
- package/lib/core/artifact-evidence.d.ts.map +1 -1
- package/lib/core/artifact-evidence.js +36 -4
- package/lib/core/artifact-evidence.js.map +1 -1
- package/lib/core/assertion.d.ts +15 -0
- package/lib/core/assertion.d.ts.map +1 -1
- package/lib/core/assertion.js +16 -0
- package/lib/core/assertion.js.map +1 -1
- package/lib/core/case.d.ts +106 -2
- package/lib/core/case.d.ts.map +1 -1
- package/lib/core/case.js +16 -2
- package/lib/core/case.js.map +1 -1
- package/lib/core/command-verification.d.ts +10 -0
- package/lib/core/command-verification.d.ts.map +1 -0
- package/lib/core/command-verification.js +66 -0
- package/lib/core/command-verification.js.map +1 -0
- package/lib/core/driver-execution.d.ts +83 -0
- package/lib/core/driver-execution.d.ts.map +1 -0
- package/lib/core/driver-execution.js +72 -0
- package/lib/core/driver-execution.js.map +1 -0
- package/lib/core/driver-metrics.d.ts +98 -0
- package/lib/core/driver-metrics.d.ts.map +1 -0
- package/lib/core/driver-metrics.js +36 -0
- package/lib/core/driver-metrics.js.map +1 -0
- package/lib/core/driver.d.ts +64 -0
- package/lib/core/driver.d.ts.map +1 -1
- package/lib/core/driver.js +28 -0
- package/lib/core/driver.js.map +1 -1
- package/lib/core/duration.d.ts +5 -0
- package/lib/core/duration.d.ts.map +1 -0
- package/lib/core/duration.js +49 -0
- package/lib/core/duration.js.map +1 -0
- package/lib/core/environment.d.ts +18 -0
- package/lib/core/environment.d.ts.map +1 -0
- package/lib/core/environment.js +21 -0
- package/lib/core/environment.js.map +1 -0
- package/lib/core/evaluation-clusters.d.ts +17 -0
- package/lib/core/evaluation-clusters.d.ts.map +1 -0
- package/lib/core/evaluation-clusters.js +23 -0
- package/lib/core/evaluation-clusters.js.map +1 -0
- package/lib/core/events.d.ts +36 -0
- package/lib/core/events.d.ts.map +1 -1
- package/lib/core/events.js +12 -2
- package/lib/core/events.js.map +1 -1
- package/lib/core/experiment.d.ts +32 -2
- package/lib/core/experiment.d.ts.map +1 -1
- package/lib/core/experiment.js +35 -18
- package/lib/core/experiment.js.map +1 -1
- package/lib/core/file-tree.d.ts +7 -0
- package/lib/core/file-tree.d.ts.map +1 -0
- package/lib/core/file-tree.js +35 -0
- package/lib/core/file-tree.js.map +1 -0
- package/lib/core/global-id.d.ts +6 -0
- package/lib/core/global-id.d.ts.map +1 -0
- package/lib/core/global-id.js +17 -0
- package/lib/core/global-id.js.map +1 -0
- package/lib/core/harness-catalog.d.ts +35 -0
- package/lib/core/harness-catalog.d.ts.map +1 -0
- package/lib/core/harness-catalog.js +45 -0
- package/lib/core/harness-catalog.js.map +1 -0
- package/lib/core/harness-catalogs/antigravity.d.ts +11 -0
- package/lib/core/harness-catalogs/antigravity.d.ts.map +1 -0
- package/lib/core/harness-catalogs/antigravity.js +23 -0
- package/lib/core/harness-catalogs/antigravity.js.map +1 -0
- package/lib/core/harness-catalogs/claude-code.d.ts +12 -0
- package/lib/core/harness-catalogs/claude-code.d.ts.map +1 -0
- package/lib/core/harness-catalogs/claude-code.js +24 -0
- package/lib/core/harness-catalogs/claude-code.js.map +1 -0
- package/lib/core/harness-catalogs/codex.d.ts +11 -0
- package/lib/core/harness-catalogs/codex.d.ts.map +1 -0
- package/lib/core/harness-catalogs/codex.js +26 -0
- package/lib/core/harness-catalogs/codex.js.map +1 -0
- package/lib/core/harness-catalogs/copilot.d.ts +29 -0
- package/lib/core/harness-catalogs/copilot.d.ts.map +1 -0
- package/lib/core/harness-catalogs/copilot.js +45 -0
- package/lib/core/harness-catalogs/copilot.js.map +1 -0
- package/lib/core/harness-catalogs/cursor.d.ts +2 -0
- package/lib/core/harness-catalogs/cursor.d.ts.map +1 -0
- package/lib/core/harness-catalogs/cursor.js +13 -0
- package/lib/core/harness-catalogs/cursor.js.map +1 -0
- package/lib/core/harness-catalogs/grok-build.d.ts +7 -0
- package/lib/core/harness-catalogs/grok-build.d.ts.map +1 -0
- package/lib/core/harness-catalogs/grok-build.js +18 -0
- package/lib/core/harness-catalogs/grok-build.js.map +1 -0
- package/lib/core/harness-catalogs/index.d.ts +73 -0
- package/lib/core/harness-catalogs/index.d.ts.map +1 -0
- package/lib/core/harness-catalogs/index.js +17 -0
- package/lib/core/harness-catalogs/index.js.map +1 -0
- package/lib/core/harness-matrix.d.ts +154 -0
- package/lib/core/harness-matrix.d.ts.map +1 -0
- package/lib/core/harness-matrix.js +204 -0
- package/lib/core/harness-matrix.js.map +1 -0
- package/lib/core/index.d.ts +12 -0
- package/lib/core/index.d.ts.map +1 -1
- package/lib/core/index.js +12 -0
- package/lib/core/index.js.map +1 -1
- package/lib/core/lifecycle-fixtures.d.ts.map +1 -1
- package/lib/core/lifecycle-fixtures.js +4 -2
- package/lib/core/lifecycle-fixtures.js.map +1 -1
- package/lib/core/model-cli.d.ts +9 -11
- package/lib/core/model-cli.d.ts.map +1 -1
- package/lib/core/model-cli.js +31 -19
- package/lib/core/model-cli.js.map +1 -1
- package/lib/core/platform-process.d.ts +6 -1
- package/lib/core/platform-process.d.ts.map +1 -1
- package/lib/core/platform-process.js +49 -11
- package/lib/core/platform-process.js.map +1 -1
- package/lib/core/run.d.ts +292 -1
- package/lib/core/run.d.ts.map +1 -1
- package/lib/core/run.js +27 -0
- package/lib/core/run.js.map +1 -1
- package/lib/core/secure-copy.d.ts.map +1 -0
- package/lib/core/secure-copy.js.map +1 -0
- package/lib/core/usage.d.ts +5 -0
- package/lib/core/usage.d.ts.map +1 -1
- package/lib/core/usage.js +3 -0
- package/lib/core/usage.js.map +1 -1
- package/lib/core/wire.d.ts +12 -0
- package/lib/core/wire.d.ts.map +1 -1
- package/lib/core/wire.js +5 -1
- package/lib/core/wire.js.map +1 -1
- package/lib/drivers/antigravity/artifacts.d.ts +2 -0
- package/lib/drivers/antigravity/artifacts.d.ts.map +1 -0
- package/lib/drivers/antigravity/artifacts.js +33 -0
- package/lib/drivers/antigravity/artifacts.js.map +1 -0
- package/lib/drivers/antigravity/config.d.ts +88 -0
- package/lib/drivers/antigravity/config.d.ts.map +1 -0
- package/lib/drivers/antigravity/config.js +87 -0
- package/lib/drivers/antigravity/config.js.map +1 -0
- package/lib/drivers/antigravity/events.d.ts +4 -0
- package/lib/drivers/antigravity/events.d.ts.map +1 -0
- package/lib/drivers/antigravity/events.js +241 -0
- package/lib/drivers/antigravity/events.js.map +1 -0
- package/lib/drivers/antigravity/index.d.ts +114 -0
- package/lib/drivers/antigravity/index.d.ts.map +1 -0
- package/lib/drivers/antigravity/index.js +113 -0
- package/lib/drivers/antigravity/index.js.map +1 -0
- package/lib/drivers/antigravity/runtime.d.ts +7 -0
- package/lib/drivers/antigravity/runtime.d.ts.map +1 -0
- package/lib/drivers/antigravity/runtime.js +83 -0
- package/lib/drivers/antigravity/runtime.js.map +1 -0
- package/lib/drivers/claude-code/deferred-process.d.ts +17 -0
- package/lib/drivers/claude-code/deferred-process.d.ts.map +1 -0
- package/lib/drivers/claude-code/deferred-process.js +108 -0
- package/lib/drivers/claude-code/deferred-process.js.map +1 -0
- package/lib/drivers/claude-code/index.d.ts +180 -10
- package/lib/drivers/claude-code/index.d.ts.map +1 -1
- package/lib/drivers/claude-code/index.js +186 -252
- package/lib/drivers/claude-code/index.js.map +1 -1
- package/lib/drivers/claude-code/isolation.d.ts +8 -0
- package/lib/drivers/claude-code/isolation.d.ts.map +1 -0
- package/lib/drivers/claude-code/isolation.js +199 -0
- package/lib/drivers/claude-code/isolation.js.map +1 -0
- package/lib/drivers/claude-code/parse-events.d.ts.map +1 -1
- package/lib/drivers/claude-code/parse-events.js +73 -8
- package/lib/drivers/claude-code/parse-events.js.map +1 -1
- package/lib/drivers/codex/app-events.d.ts +9 -2
- package/lib/drivers/codex/app-events.d.ts.map +1 -1
- package/lib/drivers/codex/app-events.js +104 -12
- package/lib/drivers/codex/app-events.js.map +1 -1
- package/lib/drivers/codex/config.d.ts +66 -5
- package/lib/drivers/codex/config.d.ts.map +1 -1
- package/lib/drivers/codex/config.js +39 -8
- package/lib/drivers/codex/config.js.map +1 -1
- package/lib/drivers/codex/exec-events.d.ts +4 -0
- package/lib/drivers/codex/exec-events.d.ts.map +1 -0
- package/lib/drivers/codex/exec-events.js +230 -0
- package/lib/drivers/codex/exec-events.js.map +1 -0
- package/lib/drivers/codex/index.d.ts +138 -4
- package/lib/drivers/codex/index.d.ts.map +1 -1
- package/lib/drivers/codex/index.js +112 -276
- package/lib/drivers/codex/index.js.map +1 -1
- package/lib/drivers/copilot/config.d.ts +84 -0
- package/lib/drivers/copilot/config.d.ts.map +1 -0
- package/lib/drivers/copilot/config.js +57 -0
- package/lib/drivers/copilot/config.js.map +1 -0
- package/lib/drivers/copilot/index.d.ts +107 -0
- package/lib/drivers/copilot/index.d.ts.map +1 -0
- package/lib/drivers/copilot/index.js +268 -0
- package/lib/drivers/copilot/index.js.map +1 -0
- package/lib/drivers/cursor/config.d.ts +28 -0
- package/lib/drivers/cursor/config.d.ts.map +1 -0
- package/lib/drivers/cursor/config.js +28 -0
- package/lib/drivers/cursor/config.js.map +1 -0
- package/lib/drivers/cursor/index.d.ts +44 -0
- package/lib/drivers/cursor/index.d.ts.map +1 -0
- package/lib/drivers/cursor/index.js +286 -0
- package/lib/drivers/cursor/index.js.map +1 -0
- package/lib/drivers/cursor/interaction.d.ts +6 -0
- package/lib/drivers/cursor/interaction.d.ts.map +1 -0
- package/lib/drivers/cursor/interaction.js +55 -0
- package/lib/drivers/cursor/interaction.js.map +1 -0
- package/lib/drivers/grok-build/config.d.ts +68 -0
- package/lib/drivers/grok-build/config.d.ts.map +1 -0
- package/lib/drivers/grok-build/config.js +40 -0
- package/lib/drivers/grok-build/config.js.map +1 -0
- package/lib/drivers/grok-build/index.d.ts +87 -0
- package/lib/drivers/grok-build/index.d.ts.map +1 -0
- package/lib/drivers/grok-build/index.js +293 -0
- package/lib/drivers/grok-build/index.js.map +1 -0
- package/lib/drivers/mcp/index.d.ts +37 -4
- package/lib/drivers/mcp/index.d.ts.map +1 -1
- package/lib/drivers/mcp/index.js +25 -11
- package/lib/drivers/mcp/index.js.map +1 -1
- package/lib/drivers/shared/acp-events.d.ts +15 -0
- package/lib/drivers/shared/acp-events.d.ts.map +1 -0
- package/lib/drivers/shared/acp-events.js +83 -0
- package/lib/drivers/shared/acp-events.js.map +1 -0
- package/lib/drivers/shared/acp-interaction.d.ts +7 -0
- package/lib/drivers/shared/acp-interaction.d.ts.map +1 -0
- package/lib/drivers/shared/acp-interaction.js +20 -0
- package/lib/drivers/shared/acp-interaction.js.map +1 -0
- package/lib/drivers/shared/acp-session.d.ts +4 -0
- package/lib/drivers/shared/acp-session.d.ts.map +1 -0
- package/lib/drivers/shared/acp-session.js +23 -0
- package/lib/drivers/shared/acp-session.js.map +1 -0
- package/lib/drivers/shared/acp-transport.d.ts +40 -0
- package/lib/drivers/shared/acp-transport.d.ts.map +1 -0
- package/lib/drivers/shared/acp-transport.js +152 -0
- package/lib/drivers/shared/acp-transport.js.map +1 -0
- package/lib/drivers/shared/artifacts.d.ts.map +1 -1
- package/lib/drivers/shared/artifacts.js +3 -2
- package/lib/drivers/shared/artifacts.js.map +1 -1
- package/lib/drivers/shared/experiment-behavior.d.ts +13 -0
- package/lib/drivers/shared/experiment-behavior.d.ts.map +1 -0
- package/lib/drivers/shared/experiment-behavior.js +197 -0
- package/lib/drivers/shared/experiment-behavior.js.map +1 -0
- package/lib/drivers/shared/experiment-policy.d.ts +4 -0
- package/lib/drivers/shared/experiment-policy.d.ts.map +1 -0
- package/lib/drivers/shared/experiment-policy.js +143 -0
- package/lib/drivers/shared/experiment-policy.js.map +1 -0
- package/lib/drivers/subprocess/index.d.ts +39 -4
- package/lib/drivers/subprocess/index.d.ts.map +1 -1
- package/lib/drivers/subprocess/index.js +56 -12
- package/lib/drivers/subprocess/index.js.map +1 -1
- package/lib/index.d.ts +22 -8
- package/lib/index.d.ts.map +1 -1
- package/lib/index.js +20 -6
- package/lib/index.js.map +1 -1
- package/lib/optimization/reflective.d.ts +3 -1
- package/lib/optimization/reflective.d.ts.map +1 -1
- package/lib/optimization/reflective.js +3 -2
- package/lib/optimization/reflective.js.map +1 -1
- package/lib/orchestrator/annotation-loader.d.ts +6 -2
- package/lib/orchestrator/annotation-loader.d.ts.map +1 -1
- package/lib/orchestrator/annotation-loader.js +8 -0
- package/lib/orchestrator/annotation-loader.js.map +1 -1
- package/lib/orchestrator/annotation-store.d.ts.map +1 -1
- package/lib/orchestrator/annotation-store.js +16 -2
- package/lib/orchestrator/annotation-store.js.map +1 -1
- package/lib/orchestrator/artifact-origin.d.ts +3 -0
- package/lib/orchestrator/artifact-origin.d.ts.map +1 -0
- package/lib/orchestrator/artifact-origin.js +40 -0
- package/lib/orchestrator/artifact-origin.js.map +1 -0
- package/lib/orchestrator/campaign-lifecycle.d.ts +1 -0
- package/lib/orchestrator/campaign-lifecycle.d.ts.map +1 -1
- package/lib/orchestrator/campaign-lifecycle.js +4 -1
- package/lib/orchestrator/campaign-lifecycle.js.map +1 -1
- package/lib/orchestrator/candidates.d.ts.map +1 -1
- package/lib/orchestrator/candidates.js +3 -2
- package/lib/orchestrator/candidates.js.map +1 -1
- package/lib/orchestrator/compare.d.ts +17 -1
- package/lib/orchestrator/compare.d.ts.map +1 -1
- package/lib/orchestrator/compare.js +25 -1
- package/lib/orchestrator/compare.js.map +1 -1
- package/lib/orchestrator/comparison-identity.d.ts.map +1 -1
- package/lib/orchestrator/comparison-identity.js +2 -0
- package/lib/orchestrator/comparison-identity.js.map +1 -1
- package/lib/orchestrator/config.d.ts +177 -1
- package/lib/orchestrator/config.d.ts.map +1 -1
- package/lib/orchestrator/config.js +18 -4
- package/lib/orchestrator/config.js.map +1 -1
- package/lib/orchestrator/discover.d.ts +1 -0
- package/lib/orchestrator/discover.d.ts.map +1 -1
- package/lib/orchestrator/discover.js +7 -1
- package/lib/orchestrator/discover.js.map +1 -1
- package/lib/orchestrator/doctor.d.ts +1 -0
- package/lib/orchestrator/doctor.d.ts.map +1 -1
- package/lib/orchestrator/doctor.js +121 -33
- package/lib/orchestrator/doctor.js.map +1 -1
- package/lib/orchestrator/driver-capabilities.d.ts +4 -0
- package/lib/orchestrator/driver-capabilities.d.ts.map +1 -0
- package/lib/orchestrator/driver-capabilities.js +13 -0
- package/lib/orchestrator/driver-capabilities.js.map +1 -0
- package/lib/orchestrator/driver-identity.d.ts +5 -0
- package/lib/orchestrator/driver-identity.d.ts.map +1 -0
- package/lib/orchestrator/driver-identity.js +51 -0
- package/lib/orchestrator/driver-identity.js.map +1 -0
- package/lib/orchestrator/experiment-context.d.ts +51 -3
- package/lib/orchestrator/experiment-context.d.ts.map +1 -1
- package/lib/orchestrator/experiment-context.js.map +1 -1
- package/lib/orchestrator/experiment-contracts.d.ts +30 -1
- package/lib/orchestrator/experiment-contracts.d.ts.map +1 -1
- package/lib/orchestrator/experiment-corpus.d.ts +2 -1
- package/lib/orchestrator/experiment-corpus.d.ts.map +1 -1
- package/lib/orchestrator/experiment-corpus.js +28 -22
- package/lib/orchestrator/experiment-corpus.js.map +1 -1
- package/lib/orchestrator/experiment-effects.d.ts +33 -0
- package/lib/orchestrator/experiment-effects.d.ts.map +1 -0
- package/lib/orchestrator/experiment-effects.js +44 -0
- package/lib/orchestrator/experiment-effects.js.map +1 -0
- package/lib/orchestrator/experiment-evaluation.d.ts.map +1 -1
- package/lib/orchestrator/experiment-evaluation.js +56 -5
- package/lib/orchestrator/experiment-evaluation.js.map +1 -1
- package/lib/orchestrator/experiment-identity.d.ts +3 -0
- package/lib/orchestrator/experiment-identity.d.ts.map +1 -1
- package/lib/orchestrator/experiment-identity.js +22 -33
- package/lib/orchestrator/experiment-identity.js.map +1 -1
- package/lib/orchestrator/experiment-loader.d.ts +8 -3
- package/lib/orchestrator/experiment-loader.d.ts.map +1 -1
- package/lib/orchestrator/experiment-preflight.d.ts +3 -0
- package/lib/orchestrator/experiment-preflight.d.ts.map +1 -0
- package/lib/orchestrator/experiment-preflight.js +21 -0
- package/lib/orchestrator/experiment-preflight.js.map +1 -0
- package/lib/orchestrator/experiment-report.d.ts +90 -1
- package/lib/orchestrator/experiment-report.d.ts.map +1 -1
- package/lib/orchestrator/experiment-report.js +9 -1
- package/lib/orchestrator/experiment-report.js.map +1 -1
- package/lib/orchestrator/experiment-runs.d.ts +61 -2
- package/lib/orchestrator/experiment-runs.d.ts.map +1 -1
- package/lib/orchestrator/experiment-runs.js +13 -5
- package/lib/orchestrator/experiment-runs.js.map +1 -1
- package/lib/orchestrator/experiment-runtime-identity.d.ts +0 -16
- package/lib/orchestrator/experiment-runtime-identity.d.ts.map +1 -1
- package/lib/orchestrator/experiment-runtime-identity.js +5 -220
- package/lib/orchestrator/experiment-runtime-identity.js.map +1 -1
- package/lib/orchestrator/experiment-selection.d.ts.map +1 -1
- package/lib/orchestrator/experiment-selection.js +52 -0
- package/lib/orchestrator/experiment-selection.js.map +1 -1
- package/lib/orchestrator/experiment.d.ts +3 -3
- package/lib/orchestrator/experiment.d.ts.map +1 -1
- package/lib/orchestrator/experiment.js +26 -21
- package/lib/orchestrator/experiment.js.map +1 -1
- package/lib/orchestrator/fixture-identity.d.ts +1 -0
- package/lib/orchestrator/fixture-identity.d.ts.map +1 -1
- package/lib/orchestrator/fixture-identity.js +6 -1
- package/lib/orchestrator/fixture-identity.js.map +1 -1
- package/lib/orchestrator/fs-read.d.ts +1 -6
- package/lib/orchestrator/fs-read.d.ts.map +1 -1
- package/lib/orchestrator/fs-read.js +2 -32
- package/lib/orchestrator/fs-read.js.map +1 -1
- package/lib/orchestrator/grader-experiment.d.ts +29 -1
- package/lib/orchestrator/grader-experiment.d.ts.map +1 -1
- package/lib/orchestrator/grader-experiment.js +54 -9
- package/lib/orchestrator/grader-experiment.js.map +1 -1
- package/lib/orchestrator/grading.d.ts +2 -1
- package/lib/orchestrator/grading.d.ts.map +1 -1
- package/lib/orchestrator/grading.js +27 -8
- package/lib/orchestrator/grading.js.map +1 -1
- package/lib/orchestrator/harness-catalog-drift.d.ts +22 -0
- package/lib/orchestrator/harness-catalog-drift.d.ts.map +1 -0
- package/lib/orchestrator/harness-catalog-drift.js +90 -0
- package/lib/orchestrator/harness-catalog-drift.js.map +1 -0
- package/lib/orchestrator/harness-registry.d.ts +4 -0
- package/lib/orchestrator/harness-registry.d.ts.map +1 -0
- package/lib/orchestrator/harness-registry.js +5 -0
- package/lib/orchestrator/harness-registry.js.map +1 -0
- package/lib/orchestrator/index.d.ts +1 -1
- package/lib/orchestrator/index.d.ts.map +1 -1
- package/lib/orchestrator/index.js +1 -1
- package/lib/orchestrator/index.js.map +1 -1
- package/lib/orchestrator/list-runs.d.ts +8 -1
- package/lib/orchestrator/list-runs.d.ts.map +1 -1
- package/lib/orchestrator/list-runs.js +77 -12
- package/lib/orchestrator/list-runs.js.map +1 -1
- package/lib/orchestrator/optimization-lifecycle.d.ts.map +1 -1
- package/lib/orchestrator/optimization-lifecycle.js +3 -1
- package/lib/orchestrator/optimization-lifecycle.js.map +1 -1
- package/lib/orchestrator/orchestrator.d.ts +18 -2
- package/lib/orchestrator/orchestrator.d.ts.map +1 -1
- package/lib/orchestrator/orchestrator.js +50 -168
- package/lib/orchestrator/orchestrator.js.map +1 -1
- package/lib/orchestrator/paired-effects.d.ts +34 -0
- package/lib/orchestrator/paired-effects.d.ts.map +1 -0
- package/lib/orchestrator/paired-effects.js +158 -0
- package/lib/orchestrator/paired-effects.js.map +1 -0
- package/lib/orchestrator/provenance.d.ts +5 -1
- package/lib/orchestrator/provenance.d.ts.map +1 -1
- package/lib/orchestrator/provenance.js +16 -3
- package/lib/orchestrator/provenance.js.map +1 -1
- package/lib/orchestrator/regrade.d.ts +7 -0
- package/lib/orchestrator/regrade.d.ts.map +1 -1
- package/lib/orchestrator/rubric-loader.js +1 -1
- package/lib/orchestrator/rubric-loader.js.map +1 -1
- package/lib/orchestrator/run-dir.d.ts +2 -0
- package/lib/orchestrator/run-dir.d.ts.map +1 -1
- package/lib/orchestrator/run-dir.js +1 -1
- package/lib/orchestrator/run-dir.js.map +1 -1
- package/lib/orchestrator/run-execution.d.ts +3 -0
- package/lib/orchestrator/run-execution.d.ts.map +1 -1
- package/lib/orchestrator/run-execution.js +66 -12
- package/lib/orchestrator/run-execution.js.map +1 -1
- package/lib/orchestrator/run-lifecycle.d.ts +1 -0
- package/lib/orchestrator/run-lifecycle.d.ts.map +1 -1
- package/lib/orchestrator/run-lifecycle.js +3 -2
- package/lib/orchestrator/run-lifecycle.js.map +1 -1
- package/lib/orchestrator/run-workspace.d.ts +10 -0
- package/lib/orchestrator/run-workspace.d.ts.map +1 -0
- package/lib/orchestrator/run-workspace.js +58 -0
- package/lib/orchestrator/run-workspace.js.map +1 -0
- package/lib/orchestrator/suite-compare.d.ts +37 -0
- package/lib/orchestrator/suite-compare.d.ts.map +1 -1
- package/lib/orchestrator/suite-compare.js +27 -17
- package/lib/orchestrator/suite-compare.js.map +1 -1
- package/lib/orchestrator/suite-runs.d.ts +6 -1
- package/lib/orchestrator/suite-runs.d.ts.map +1 -1
- package/lib/orchestrator/suite-runs.js +14 -2
- package/lib/orchestrator/suite-runs.js.map +1 -1
- package/lib/orchestrator/suite-summary.d.ts +77 -0
- package/lib/orchestrator/suite-summary.d.ts.map +1 -1
- package/lib/orchestrator/suite-summary.js +61 -12
- package/lib/orchestrator/suite-summary.js.map +1 -1
- package/lib/previewer/artifacts.d.ts +25 -0
- package/lib/previewer/artifacts.d.ts.map +1 -0
- package/lib/previewer/artifacts.js +180 -0
- package/lib/previewer/artifacts.js.map +1 -0
- package/lib/previewer/server.d.ts +111 -0
- package/lib/previewer/server.d.ts.map +1 -0
- package/lib/previewer/server.js +558 -0
- package/lib/previewer/server.js.map +1 -0
- package/package.json +3 -3
- package/skills/lux-answerer/SKILL.md +2 -2
- package/src/answerers/claude-code/index.ts +6 -5
- package/src/answerers/persona/index.ts +75 -34
- package/src/answerers/terminal/index.ts +8 -12
- package/src/assertions/authoring.ts +2 -0
- package/src/assertions/core/command-exits-zero.ts +61 -7
- package/src/assertions/core/event-checks.ts +64 -3
- package/src/assertions/core/helpers.ts +3 -6
- package/src/assertions/core/index.ts +2 -0
- package/src/assertions/core/resource-checks.ts +38 -5
- package/src/assertions/rubric/index.ts +68 -13
- package/src/assertions/rubric/internal.ts +227 -27
- package/src/authoring.ts +345 -2
- package/src/cli/bin.ts +2 -12
- package/src/cli/command-runtime.ts +20 -9
- package/src/cli/doctor.ts +17 -0
- package/src/cli/harness-plan.ts +117 -0
- package/src/cli/init-templates.ts +238 -0
- package/src/cli/init.ts +53 -184
- package/src/cli/output-schemas.ts +106 -0
- package/src/cli/program.ts +689 -721
- package/src/cli/render/reporter.ts +43 -7
- package/src/cli/run-options.ts +132 -25
- package/src/cli/view.ts +129 -0
- package/src/container/driver.ts +211 -0
- package/src/container/protocol.ts +48 -0
- package/src/container/resources.ts +153 -0
- package/src/container/worker.ts +107 -0
- package/src/core/action-events.ts +59 -0
- package/src/core/agent-activity.ts +236 -0
- package/src/core/agent-environment.ts +64 -0
- package/src/core/agent-events.ts +17 -0
- package/src/core/annotation.ts +2 -1
- package/src/core/answerer.ts +9 -0
- package/src/core/artifact-evidence.ts +47 -4
- package/src/core/assertion.ts +18 -0
- package/src/core/case.ts +16 -2
- package/src/core/command-verification.ts +74 -0
- package/src/core/driver-execution.ts +84 -0
- package/src/core/driver-metrics.ts +51 -0
- package/src/core/driver.ts +76 -0
- package/src/core/duration.ts +52 -0
- package/src/core/environment.ts +22 -0
- package/src/core/evaluation-clusters.ts +35 -0
- package/src/core/events.ts +14 -2
- package/src/core/experiment.ts +64 -24
- package/src/core/file-tree.ts +35 -0
- package/src/core/global-id.ts +26 -0
- package/src/core/harness-catalog.ts +100 -0
- package/src/core/harness-catalogs/antigravity.ts +24 -0
- package/src/core/harness-catalogs/claude-code.ts +25 -0
- package/src/core/harness-catalogs/codex.ts +27 -0
- package/src/core/harness-catalogs/copilot.ts +46 -0
- package/src/core/harness-catalogs/cursor.ts +13 -0
- package/src/core/harness-catalogs/grok-build.ts +19 -0
- package/src/core/harness-catalogs/index.ts +31 -0
- package/src/core/harness-matrix.ts +316 -0
- package/src/core/index.ts +12 -0
- package/src/core/lifecycle-fixtures.ts +4 -2
- package/src/core/model-cli.ts +52 -23
- package/src/core/platform-process.ts +58 -10
- package/src/core/run.ts +31 -0
- package/src/core/usage.ts +8 -0
- package/src/core/wire.ts +5 -1
- package/src/drivers/antigravity/README.md +171 -0
- package/src/drivers/antigravity/artifacts.ts +37 -0
- package/src/drivers/antigravity/config.ts +92 -0
- package/src/drivers/antigravity/events.ts +290 -0
- package/src/drivers/antigravity/index.ts +132 -0
- package/src/drivers/antigravity/runtime.ts +94 -0
- package/src/drivers/claude-code/deferred-process.ts +118 -0
- package/src/drivers/claude-code/index.ts +216 -281
- package/src/drivers/claude-code/isolation.ts +228 -0
- package/src/drivers/claude-code/parse-events.ts +98 -12
- package/src/drivers/codex/app-events.ts +126 -16
- package/src/drivers/codex/config.ts +50 -8
- package/src/drivers/codex/exec-events.ts +258 -0
- package/src/drivers/codex/index.ts +144 -312
- package/src/drivers/copilot/README.md +60 -0
- package/src/drivers/copilot/config.ts +61 -0
- package/src/drivers/copilot/index.ts +310 -0
- package/src/drivers/cursor/README.md +66 -0
- package/src/drivers/cursor/config.ts +31 -0
- package/src/drivers/cursor/index.ts +326 -0
- package/src/drivers/cursor/interaction.ts +74 -0
- package/src/drivers/grok-build/README.md +112 -0
- package/src/drivers/grok-build/config.ts +43 -0
- package/src/drivers/grok-build/index.ts +331 -0
- package/src/drivers/mcp/index.ts +25 -11
- package/src/drivers/shared/acp-events.ts +94 -0
- package/src/drivers/shared/acp-interaction.ts +27 -0
- package/src/drivers/shared/acp-session.ts +40 -0
- package/src/drivers/shared/acp-transport.ts +187 -0
- package/src/drivers/shared/artifacts.ts +6 -2
- package/src/drivers/shared/experiment-behavior.ts +214 -0
- package/src/drivers/shared/experiment-policy.ts +195 -0
- package/src/drivers/subprocess/index.ts +55 -12
- package/src/index.ts +115 -10
- package/src/optimization/reflective.ts +3 -2
- package/src/orchestrator/annotation-loader.ts +10 -0
- package/src/orchestrator/annotation-store.ts +24 -3
- package/src/orchestrator/artifact-origin.ts +50 -0
- package/src/orchestrator/campaign-lifecycle.ts +5 -1
- package/src/orchestrator/candidates.ts +3 -2
- package/src/orchestrator/compare.ts +34 -1
- package/src/orchestrator/comparison-identity.ts +2 -0
- package/src/orchestrator/config.ts +20 -3
- package/src/orchestrator/discover.ts +8 -1
- package/src/orchestrator/doctor.ts +189 -28
- package/src/orchestrator/driver-capabilities.ts +21 -0
- package/src/orchestrator/driver-identity.ts +69 -0
- package/src/orchestrator/experiment-context.ts +3 -2
- package/src/orchestrator/experiment-corpus.ts +53 -37
- package/src/orchestrator/experiment-effects.ts +61 -0
- package/src/orchestrator/experiment-evaluation.ts +71 -10
- package/src/orchestrator/experiment-identity.ts +42 -48
- package/src/orchestrator/experiment-preflight.ts +27 -0
- package/src/orchestrator/experiment-report.ts +21 -1
- package/src/orchestrator/experiment-runs.ts +11 -4
- package/src/orchestrator/experiment-runtime-identity.ts +5 -243
- package/src/orchestrator/experiment-selection.ts +77 -0
- package/src/orchestrator/experiment.ts +29 -29
- package/src/orchestrator/fixture-identity.ts +9 -1
- package/src/orchestrator/fs-read.ts +2 -32
- package/src/orchestrator/grader-experiment.ts +74 -13
- package/src/orchestrator/grading.ts +35 -7
- package/src/orchestrator/harness-catalog-drift.ts +132 -0
- package/src/orchestrator/harness-registry.ts +12 -0
- package/src/orchestrator/index.ts +1 -0
- package/src/orchestrator/list-runs.ts +95 -11
- package/src/orchestrator/optimization-lifecycle.ts +4 -1
- package/src/orchestrator/orchestrator.ts +92 -216
- package/src/orchestrator/paired-effects.ts +209 -0
- package/src/orchestrator/provenance.ts +23 -3
- package/src/orchestrator/rubric-loader.ts +1 -1
- package/src/orchestrator/run-dir.ts +8 -1
- package/src/orchestrator/run-execution.ts +77 -11
- package/src/orchestrator/run-lifecycle.ts +4 -2
- package/src/orchestrator/run-workspace.ts +62 -0
- package/src/orchestrator/suite-compare.ts +29 -19
- package/src/orchestrator/suite-runs.ts +13 -2
- package/src/orchestrator/suite-summary.ts +122 -13
- package/src/previewer/artifacts.ts +221 -0
- package/src/previewer/server.ts +787 -0
- package/viewer/app.css +2075 -0
- package/viewer/app.js +627 -0
- package/viewer/artifact-view.js +48 -0
- package/viewer/charts.js +86 -0
- package/viewer/detail.js +1231 -0
- package/viewer/format.js +129 -0
- package/viewer/index.html +151 -0
- package/lib/answerers/scripted/index.d.ts +0 -8
- package/lib/answerers/scripted/index.d.ts.map +0 -1
- package/lib/answerers/scripted/index.js +0 -163
- package/lib/answerers/scripted/index.js.map +0 -1
- package/lib/cli/skills.d.ts +0 -38
- package/lib/cli/skills.d.ts.map +0 -1
- package/lib/cli/skills.js +0 -84
- package/lib/cli/skills.js.map +0 -1
- package/lib/drivers/shared/secure-copy.d.ts.map +0 -1
- package/lib/drivers/shared/secure-copy.js.map +0 -1
- package/skills/lux/SKILL.md +0 -85
- package/skills/lux/references/cli.md +0 -102
- package/skills/lux/references/eval-authoring.md +0 -105
- package/src/answerers/scripted/index.ts +0 -186
- package/src/cli/skills.ts +0 -129
- /package/lib/{drivers/shared → core}/secure-copy.d.ts +0 -0
- /package/lib/{drivers/shared → core}/secure-copy.js +0 -0
- /package/src/{drivers/shared → core}/secure-copy.ts +0 -0
|
@@ -4,6 +4,7 @@ import {
|
|
|
4
4
|
createToolOutcomeAccumulator,
|
|
5
5
|
emptyRunUsage,
|
|
6
6
|
type GradingAssertionEntry,
|
|
7
|
+
type GradingReport,
|
|
7
8
|
harnessUsage,
|
|
8
9
|
isUsageEmpty,
|
|
9
10
|
type LifecycleStage,
|
|
@@ -18,6 +19,7 @@ import {
|
|
|
18
19
|
TurnUsageSchema,
|
|
19
20
|
totalTokens,
|
|
20
21
|
} from "../../core/index.js";
|
|
22
|
+
import { hasAgentUsage } from "../../core/usage.js";
|
|
21
23
|
import type { CaseObserver, SuiteCaseResult, SuiteObserver } from "../../orchestrator/index.js";
|
|
22
24
|
import {
|
|
23
25
|
formatAlignedUsage,
|
|
@@ -77,7 +79,9 @@ type CaseView = {
|
|
|
77
79
|
assertionTotal: number;
|
|
78
80
|
currentAssertion: string | null;
|
|
79
81
|
assertions: GradingAssertionEntry[];
|
|
82
|
+
skippedAssertions: NonNullable<GradingReport["skippedAssertions"]>;
|
|
80
83
|
usage: RunUsage;
|
|
84
|
+
agentUsageReported: boolean;
|
|
81
85
|
/** Per-API-call agent usage, in order — the split behind `usage.agent`. */
|
|
82
86
|
agentTurns: TurnUsage[];
|
|
83
87
|
phaseDurationMs: PhaseDurations | null;
|
|
@@ -117,7 +121,10 @@ const statusMark = (status: CaseStatus, theme: Theme): string => {
|
|
|
117
121
|
|
|
118
122
|
const assertionTally = (view: CaseView): string => {
|
|
119
123
|
const passed = view.assertions.filter((a) => a.passed).length;
|
|
120
|
-
|
|
124
|
+
const evaluated = view.assertions.length === 0 ? "—" : `${passed}/${view.assertions.length}`;
|
|
125
|
+
return view.skippedAssertions.length > 0
|
|
126
|
+
? `${evaluated} · ${view.skippedAssertions.length} skipped`
|
|
127
|
+
: evaluated;
|
|
121
128
|
};
|
|
122
129
|
|
|
123
130
|
/**
|
|
@@ -154,6 +161,7 @@ const cell = (text: string, width: number): string => {
|
|
|
154
161
|
*/
|
|
155
162
|
const scoreNote = (entry: GradingAssertionEntry): string => {
|
|
156
163
|
if (entry.errorMessage) return entry.errorMessage;
|
|
164
|
+
if (entry.abstention) return "insufficient evidence";
|
|
157
165
|
if (entry.passed) return "";
|
|
158
166
|
const score = formatScore(entry.score);
|
|
159
167
|
if (entry.threshold > 0) return `score ${score} < ${formatScore(entry.threshold)}`;
|
|
@@ -233,7 +241,9 @@ export class ConsoleReporter implements SuiteObserver {
|
|
|
233
241
|
assertionTotal: 0,
|
|
234
242
|
currentAssertion: null,
|
|
235
243
|
assertions: [],
|
|
244
|
+
skippedAssertions: [],
|
|
236
245
|
usage: emptyRunUsage(),
|
|
246
|
+
agentUsageReported: false,
|
|
237
247
|
agentTurns: [],
|
|
238
248
|
phaseDurationMs: null,
|
|
239
249
|
runDir: null,
|
|
@@ -267,6 +277,13 @@ export class ConsoleReporter implements SuiteObserver {
|
|
|
267
277
|
view.toolUnresolved = tools.unresolved;
|
|
268
278
|
view.toolUnmatchedResults = tools.unmatchedResults;
|
|
269
279
|
view.toolMalformed = tools.malformedEvents;
|
|
280
|
+
} else if (event.type === "progress" && event.kind === "driver-stderr" && this.verbose) {
|
|
281
|
+
const payload = event.payload as { text?: unknown; truncated?: unknown } | null;
|
|
282
|
+
if (payload && typeof payload.text === "string") {
|
|
283
|
+
this.region.staticLine(this.theme.dim(` [${view.caseId}] ${payload.text}`));
|
|
284
|
+
if (payload.truncated === true)
|
|
285
|
+
this.region.staticLine(this.theme.dim(" [driver stderr truncated]"));
|
|
286
|
+
}
|
|
270
287
|
} else if (event.type === "progress" && event.kind === "usage") {
|
|
271
288
|
const turn = TurnUsageSchema.safeParse(event.payload).data;
|
|
272
289
|
if (turn) view.agentTurns.push(turn);
|
|
@@ -299,7 +316,9 @@ export class ConsoleReporter implements SuiteObserver {
|
|
|
299
316
|
if (result.errorMessage) view.errorMessage = result.errorMessage;
|
|
300
317
|
if (result.result) {
|
|
301
318
|
view.assertions = result.result.grading.assertions;
|
|
319
|
+
view.skippedAssertions = result.result.grading.skippedAssertions ?? [];
|
|
302
320
|
view.usage = result.result.metadata.usage;
|
|
321
|
+
view.agentUsageReported = hasAgentUsage(result.result.metadata);
|
|
303
322
|
view.phaseDurationMs = result.result.metadata.phaseDurationMs ?? null;
|
|
304
323
|
const tools = result.result.metadata.toolOutcomes;
|
|
305
324
|
if (tools) {
|
|
@@ -315,6 +334,7 @@ export class ConsoleReporter implements SuiteObserver {
|
|
|
315
334
|
// Assertions reported before the run failed to persist are not results
|
|
316
335
|
// and must not inflate the closing tally of a case that errored.
|
|
317
336
|
view.assertions = [];
|
|
337
|
+
view.skippedAssertions = [];
|
|
318
338
|
}
|
|
319
339
|
|
|
320
340
|
this.region.staticLine(this.finishedLine(view));
|
|
@@ -443,15 +463,23 @@ export class ConsoleReporter implements SuiteObserver {
|
|
|
443
463
|
Math.max(NAME_MIN, ...view.assertions.map((a) => a.name.length)),
|
|
444
464
|
);
|
|
445
465
|
for (const entry of view.assertions) {
|
|
446
|
-
|
|
466
|
+
let mark = entry.passed ? this.theme.green("✓") : this.theme.red("✗");
|
|
467
|
+
if (entry.abstention) mark = this.theme.dim("?");
|
|
447
468
|
const duration = formatDuration(entry.durationMs).padStart(DURATION_COL);
|
|
448
469
|
const note = scoreNote(entry);
|
|
449
|
-
const suffix = note
|
|
470
|
+
const suffix = note
|
|
471
|
+
? ` ${entry.abstention ? this.theme.dim(note) : this.theme.red(note)}`
|
|
472
|
+
: "";
|
|
450
473
|
nodes.push({
|
|
451
474
|
head: `${mark} ${cell(entry.name, width)} ${this.theme.dim(duration)}${suffix}`,
|
|
452
475
|
body: !entry.passed && entry.comment ? [this.theme.dim(entry.comment)] : [],
|
|
453
476
|
});
|
|
454
477
|
}
|
|
478
|
+
for (const skipped of view.skippedAssertions) {
|
|
479
|
+
nodes.push({
|
|
480
|
+
head: `${this.theme.dim("⊘")} ${skipped.name} ${this.theme.dim("skipped: judge disabled")}`,
|
|
481
|
+
});
|
|
482
|
+
}
|
|
455
483
|
|
|
456
484
|
if (view.toolCalls > 0 || view.toolUnmatchedResults > 0 || view.toolMalformed > 0) {
|
|
457
485
|
const parts = [
|
|
@@ -550,15 +578,17 @@ export class ConsoleReporter implements SuiteObserver {
|
|
|
550
578
|
}
|
|
551
579
|
|
|
552
580
|
const assertions = this.cases.flatMap((c) => c.assertions);
|
|
553
|
-
|
|
581
|
+
const skippedAssertions = this.cases.flatMap((c) => c.skippedAssertions);
|
|
582
|
+
if (assertions.length > 0 || skippedAssertions.length > 0) {
|
|
554
583
|
const ok = assertions.filter((a) => a.passed).length;
|
|
555
584
|
const bad = assertions.length - ok;
|
|
556
585
|
const tally = formatTally(
|
|
557
586
|
[
|
|
558
587
|
[ok, this.theme.green("passed")],
|
|
559
588
|
[bad, this.theme.red("failed")],
|
|
589
|
+
[skippedAssertions.length, this.theme.dim("skipped")],
|
|
560
590
|
],
|
|
561
|
-
assertions.length,
|
|
591
|
+
assertions.length + skippedAssertions.length,
|
|
562
592
|
);
|
|
563
593
|
lines.push(` ${label("Assertions")}${tally}`);
|
|
564
594
|
}
|
|
@@ -615,11 +645,17 @@ export class ConsoleReporter implements SuiteObserver {
|
|
|
615
645
|
emptyRunUsage(),
|
|
616
646
|
);
|
|
617
647
|
const harness = harnessUsage(total);
|
|
618
|
-
|
|
648
|
+
const reported = this.cases.filter(
|
|
649
|
+
(view) => view.agentUsageReported || !isUsageEmpty(this.agentUsageOf(view)),
|
|
650
|
+
).length;
|
|
651
|
+
if (reported === 0 && isUsageEmpty(total.agent) && isUsageEmpty(harness)) {
|
|
619
652
|
return live ? [] : [` ${label("Tokens")}${this.theme.dim("none reported")}`];
|
|
620
653
|
}
|
|
621
654
|
|
|
622
|
-
const
|
|
655
|
+
const missing = this.cases.length - reported;
|
|
656
|
+
const partial = missing > 0 ? ` (${missing} run${missing === 1 ? "" : "s"} unavailable)` : "";
|
|
657
|
+
const agentText =
|
|
658
|
+
reported === 0 ? this.theme.dim("unavailable") : `${formatUsage(total.agent)}${partial}`;
|
|
623
659
|
if (live) {
|
|
624
660
|
const harnessText = isUsageEmpty(harness)
|
|
625
661
|
? ""
|
package/src/cli/run-options.ts
CHANGED
|
@@ -1,4 +1,9 @@
|
|
|
1
|
-
import {
|
|
1
|
+
import { existsSync } from "node:fs";
|
|
2
|
+
import { basename, extname, resolve } from "node:path";
|
|
3
|
+
import { Errors } from "incur";
|
|
4
|
+
import type { HarnessFocus, ReasoningEffort } from "../core/index.js";
|
|
5
|
+
import { type DiscoveredCase, type SuiteItem, selectCases } from "../orchestrator/index.js";
|
|
6
|
+
import { planHarnessItems } from "./harness-plan.js";
|
|
2
7
|
|
|
3
8
|
export type RunOptions = {
|
|
4
9
|
tag?: string[];
|
|
@@ -8,28 +13,24 @@ export type RunOptions = {
|
|
|
8
13
|
listTags?: boolean;
|
|
9
14
|
interactive?: boolean;
|
|
10
15
|
claudeAnswerer?: boolean;
|
|
16
|
+
skipJudge?: boolean;
|
|
11
17
|
runsRoot?: string;
|
|
12
18
|
profile?: string;
|
|
19
|
+
driver?: string;
|
|
20
|
+
model?: string;
|
|
21
|
+
effort?: ReasoningEffort;
|
|
13
22
|
subjectRoot?: string;
|
|
14
23
|
verbose?: boolean;
|
|
15
24
|
};
|
|
16
25
|
|
|
17
|
-
export const collectTags = (value: string, previous: string[]): string[] =>
|
|
18
|
-
previous.concat(
|
|
19
|
-
value
|
|
20
|
-
.split(",")
|
|
21
|
-
.map((tag) => tag.trim())
|
|
22
|
-
.filter(Boolean),
|
|
23
|
-
);
|
|
24
|
-
|
|
25
|
-
export const collectValues = (value: string, previous: string[]): string[] => [...previous, value];
|
|
26
|
-
|
|
27
26
|
export const parseAnnotationAssertion = (
|
|
28
27
|
value: string,
|
|
29
28
|
): [string, { passed: boolean; score?: number }] => {
|
|
30
29
|
const match = /^(.+)=(pass|fail)(?::(.+))?$/.exec(value);
|
|
31
|
-
if (!match?.[1] || !match[2]) {
|
|
32
|
-
throw new
|
|
30
|
+
if (!match?.[1]?.trim() || !match[2]) {
|
|
31
|
+
throw new Errors.ParseError({
|
|
32
|
+
message: "--assertion must use id=pass, id=fail, or id=pass:0.8 syntax",
|
|
33
|
+
});
|
|
33
34
|
}
|
|
34
35
|
const score = match[3] === undefined ? undefined : parseUnitInterval(match[3]);
|
|
35
36
|
return [
|
|
@@ -41,19 +42,10 @@ export const parseAnnotationAssertion = (
|
|
|
41
42
|
];
|
|
42
43
|
};
|
|
43
44
|
|
|
44
|
-
|
|
45
|
-
// Number() rejects partial numbers such as "3abc" and "2.9".
|
|
45
|
+
const parseUnitInterval = (value: string): number => {
|
|
46
46
|
const parsed = Number(value);
|
|
47
|
-
if (!Number.
|
|
48
|
-
throw new
|
|
49
|
-
}
|
|
50
|
-
return parsed;
|
|
51
|
-
};
|
|
52
|
-
|
|
53
|
-
export const parseUnitInterval = (value: string): number => {
|
|
54
|
-
const parsed = Number(value);
|
|
55
|
-
if (!Number.isFinite(parsed) || parsed < 0 || parsed > 1) {
|
|
56
|
-
throw new InvalidArgumentError("must be a number between 0 and 1");
|
|
47
|
+
if (!value.trim() || !Number.isFinite(parsed) || parsed < 0 || parsed > 1) {
|
|
48
|
+
throw new Errors.ParseError({ message: "must be a number between 0 and 1" });
|
|
57
49
|
}
|
|
58
50
|
return parsed;
|
|
59
51
|
};
|
|
@@ -80,3 +72,118 @@ export const resolveAnswererOverride = (options: RunOptions): string | undefined
|
|
|
80
72
|
if (options.claudeAnswerer) return "claude-code";
|
|
81
73
|
return undefined;
|
|
82
74
|
};
|
|
75
|
+
|
|
76
|
+
const caseIdOf = (casePath: string): string => basename(casePath, extname(casePath));
|
|
77
|
+
|
|
78
|
+
export const suiteTitle = (
|
|
79
|
+
patterns: string[],
|
|
80
|
+
tags: string[],
|
|
81
|
+
explicit: SuiteItem[],
|
|
82
|
+
loop: number,
|
|
83
|
+
focus: HarnessFocus,
|
|
84
|
+
): string => {
|
|
85
|
+
const parts = [...patterns, ...explicit.map((e) => e.id), ...tags.map((t) => `#${t}`)];
|
|
86
|
+
const filter = parts.length > 0 ? ` ${parts.join(" ")}` : "";
|
|
87
|
+
const focusParts = [focus.driver, focus.model, focus.effort].filter(
|
|
88
|
+
(part): part is string => part !== undefined,
|
|
89
|
+
);
|
|
90
|
+
const focused = focusParts.length > 0 ? ` @${focusParts.join(":")}` : "";
|
|
91
|
+
const repeat = loop > 1 ? ` ×${loop}` : "";
|
|
92
|
+
return `run${filter}${focused}${repeat}`;
|
|
93
|
+
};
|
|
94
|
+
|
|
95
|
+
type FilterPartition = { explicit: SuiteItem[]; patterns: string[] };
|
|
96
|
+
|
|
97
|
+
// A filter that names an existing `.ts` file runs that file directly — it may
|
|
98
|
+
// sit outside the discovery root. Everything else is a substring pattern — but
|
|
99
|
+
// a `.ts` filter that resolves to nothing is almost always a typo'd path, so
|
|
100
|
+
// the silent downgrade to a pattern gets called out.
|
|
101
|
+
export const partitionFilters = (filters: string[], cwd: string): FilterPartition => {
|
|
102
|
+
const explicit: SuiteItem[] = [];
|
|
103
|
+
const patterns: string[] = [];
|
|
104
|
+
for (const f of filters) {
|
|
105
|
+
const abs = resolve(cwd, f);
|
|
106
|
+
if (f.endsWith(".ts") && existsSync(abs)) {
|
|
107
|
+
explicit.push({ path: abs, id: caseIdOf(f) });
|
|
108
|
+
} else {
|
|
109
|
+
if (f.endsWith(".ts")) {
|
|
110
|
+
process.stderr.write(`lux: ${abs} does not exist; treating "${f}" as a pattern\n`);
|
|
111
|
+
}
|
|
112
|
+
patterns.push(f);
|
|
113
|
+
}
|
|
114
|
+
}
|
|
115
|
+
return { explicit, patterns };
|
|
116
|
+
};
|
|
117
|
+
|
|
118
|
+
// The selected set — substring/tag matches plus explicitly named files — is
|
|
119
|
+
// the same whether it is run, `--list`ed, or `--list-tags`ed. No filters at
|
|
120
|
+
// all means "everything discovered"; any filter narrows.
|
|
121
|
+
export const selectSuite = (
|
|
122
|
+
cases: DiscoveredCase[],
|
|
123
|
+
{ explicit, patterns }: FilterPartition,
|
|
124
|
+
tags: string[],
|
|
125
|
+
): DiscoveredCase[] => {
|
|
126
|
+
const hasFilter = explicit.length > 0 || patterns.length > 0 || tags.length > 0;
|
|
127
|
+
const byPath = new Map<string, DiscoveredCase>();
|
|
128
|
+
if (!hasFilter) {
|
|
129
|
+
for (const c of cases) byPath.set(c.path, c);
|
|
130
|
+
} else {
|
|
131
|
+
if (patterns.length > 0 || tags.length > 0) {
|
|
132
|
+
for (const c of selectCases(cases, { patterns, tags })) byPath.set(c.path, c);
|
|
133
|
+
}
|
|
134
|
+
for (const e of explicit) {
|
|
135
|
+
const discovered = cases.find((c) => c.path === e.path);
|
|
136
|
+
// Discovery knows the case's declared id; the filename-derived one is
|
|
137
|
+
// only a fallback for files discovery never saw.
|
|
138
|
+
byPath.set(e.path, {
|
|
139
|
+
path: e.path,
|
|
140
|
+
id: discovered?.id ?? e.id,
|
|
141
|
+
relPath: discovered?.relPath ?? e.id,
|
|
142
|
+
tags: discovered?.tags ?? [],
|
|
143
|
+
});
|
|
144
|
+
}
|
|
145
|
+
}
|
|
146
|
+
return [...byPath.values()].sort((a, b) => a.id.localeCompare(b.id));
|
|
147
|
+
};
|
|
148
|
+
|
|
149
|
+
// `--loop` repeats each eval before moving to the next; each repeat gets a
|
|
150
|
+
// `#n` display suffix so the reporter keeps them apart, while `caseId` keeps
|
|
151
|
+
// every repeat filed under one run history.
|
|
152
|
+
export const expandLoop = (base: SuiteItem[], loop: number): SuiteItem[] =>
|
|
153
|
+
loop > 1
|
|
154
|
+
? base.flatMap((b) =>
|
|
155
|
+
Array.from({ length: loop }, (_, i) => ({
|
|
156
|
+
...b,
|
|
157
|
+
id: `${b.id} #${i + 1}`,
|
|
158
|
+
caseId: b.caseId ?? b.id,
|
|
159
|
+
})),
|
|
160
|
+
)
|
|
161
|
+
: base;
|
|
162
|
+
|
|
163
|
+
export const planSuiteItems = async (
|
|
164
|
+
selected: readonly { path: string; id: string }[],
|
|
165
|
+
options: Parameters<typeof planHarnessItems>[1],
|
|
166
|
+
): Promise<SuiteItem[]> => {
|
|
167
|
+
const plan = await planHarnessItems(selected, options);
|
|
168
|
+
for (const skipped of plan.unavailable) {
|
|
169
|
+
process.stderr.write(
|
|
170
|
+
`lux: skipping ${skipped.id}: harness command not found: ${skipped.commands.join(", ")}\n`,
|
|
171
|
+
);
|
|
172
|
+
}
|
|
173
|
+
if (plan.excluded.length > 0) {
|
|
174
|
+
process.stderr.write(
|
|
175
|
+
`lux: the harness focus excludes ${plan.excluded.length} pinned case${plan.excluded.length === 1 ? "" : "s"}: ${plan.excluded.join(", ")}\n`,
|
|
176
|
+
);
|
|
177
|
+
}
|
|
178
|
+
if (plan.items.length === 0) {
|
|
179
|
+
throw new Errors.IncurError({
|
|
180
|
+
code: plan.unavailable.length > 0 ? "NO_RUNNABLE_HARNESSES" : "NO_CASES_MATCHED",
|
|
181
|
+
message:
|
|
182
|
+
plan.unavailable.length > 0
|
|
183
|
+
? "No cases remain: all selected matrix harnesses are unavailable."
|
|
184
|
+
: "No cases remain after applying the harness focus.",
|
|
185
|
+
exitCode: 1,
|
|
186
|
+
});
|
|
187
|
+
}
|
|
188
|
+
return plan.items;
|
|
189
|
+
};
|
package/src/cli/view.ts
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
1
|
+
import { resolve } from "node:path";
|
|
2
|
+
import { Errors } from "incur";
|
|
3
|
+
import {
|
|
4
|
+
isExperimentRun,
|
|
5
|
+
listExperimentRuns,
|
|
6
|
+
listRuns,
|
|
7
|
+
listSuiteRuns,
|
|
8
|
+
loadExperimentRun,
|
|
9
|
+
loadLuxConfig,
|
|
10
|
+
loadRunDetail,
|
|
11
|
+
loadSuiteRun,
|
|
12
|
+
} from "../orchestrator/index.js";
|
|
13
|
+
import { openPreviewUrl, startPreviewServer } from "../previewer/server.js";
|
|
14
|
+
import { abortOnInterrupt, ensureDirectory, failWith } from "./command-runtime.js";
|
|
15
|
+
|
|
16
|
+
const SIGINT_EXIT_CODE = 130;
|
|
17
|
+
|
|
18
|
+
export type ViewOptions = {
|
|
19
|
+
runsRoot?: string;
|
|
20
|
+
case?: string;
|
|
21
|
+
suites?: boolean;
|
|
22
|
+
experiments?: boolean;
|
|
23
|
+
web?: boolean;
|
|
24
|
+
port?: number;
|
|
25
|
+
open?: boolean;
|
|
26
|
+
};
|
|
27
|
+
|
|
28
|
+
const previewKind = async (
|
|
29
|
+
initialPath: string | undefined,
|
|
30
|
+
options: ViewOptions,
|
|
31
|
+
): Promise<"run" | "suite" | "experiment" | undefined> => {
|
|
32
|
+
if (initialPath) {
|
|
33
|
+
if (isExperimentRun(initialPath)) return "experiment";
|
|
34
|
+
if (await loadSuiteRun(initialPath)) return "suite";
|
|
35
|
+
return "run";
|
|
36
|
+
}
|
|
37
|
+
if (options.experiments) return "experiment";
|
|
38
|
+
if (options.suites) return "suite";
|
|
39
|
+
return undefined;
|
|
40
|
+
};
|
|
41
|
+
|
|
42
|
+
export async function* webViewCommand(
|
|
43
|
+
runDirArg: string | undefined,
|
|
44
|
+
options: ViewOptions,
|
|
45
|
+
): AsyncGenerator<{ url: string; runsRoot: string }, void> {
|
|
46
|
+
const config = await loadLuxConfig(process.cwd());
|
|
47
|
+
const runsRoot = options.runsRoot ? resolve(options.runsRoot) : config.runsRoot;
|
|
48
|
+
const initialPath = runDirArg ? resolve(runDirArg) : undefined;
|
|
49
|
+
if (initialPath && !(await ensureDirectory("view", initialPath))) failWith(1);
|
|
50
|
+
const initialKind = await previewKind(initialPath, options);
|
|
51
|
+
using controller = abortOnInterrupt();
|
|
52
|
+
const preview = await startPreviewServer({
|
|
53
|
+
runsRoot,
|
|
54
|
+
...(options.port !== undefined ? { port: options.port } : {}),
|
|
55
|
+
...(initialPath ? { initialPath } : {}),
|
|
56
|
+
...(initialKind ? { initialKind } : {}),
|
|
57
|
+
...(options.case ? { caseFilter: options.case } : {}),
|
|
58
|
+
signal: controller.signal,
|
|
59
|
+
});
|
|
60
|
+
try {
|
|
61
|
+
yield { url: preview.url, runsRoot };
|
|
62
|
+
if (options.open !== false) {
|
|
63
|
+
try {
|
|
64
|
+
await openPreviewUrl(preview.url);
|
|
65
|
+
} catch (error) {
|
|
66
|
+
const message = error instanceof Error ? error.message : String(error);
|
|
67
|
+
process.stderr.write(`lux view: could not open a browser (${message})\n`);
|
|
68
|
+
}
|
|
69
|
+
}
|
|
70
|
+
await preview.closed;
|
|
71
|
+
if (controller.signal.aborted) failWith(SIGINT_EXIT_CODE);
|
|
72
|
+
} finally {
|
|
73
|
+
await preview.close();
|
|
74
|
+
}
|
|
75
|
+
}
|
|
76
|
+
|
|
77
|
+
export const viewCommand = async (
|
|
78
|
+
runDirArg: string | undefined,
|
|
79
|
+
options: ViewOptions,
|
|
80
|
+
): Promise<unknown> => {
|
|
81
|
+
if (options.port !== undefined || options.open === false) {
|
|
82
|
+
throw new Errors.ParseError({ message: "--port and --no-open require --web" });
|
|
83
|
+
}
|
|
84
|
+
if (runDirArg) {
|
|
85
|
+
const path = resolve(runDirArg);
|
|
86
|
+
if (!(await ensureDirectory("view", path))) return;
|
|
87
|
+
if (isExperimentRun(path)) {
|
|
88
|
+
const experiment = await loadExperimentRun(path);
|
|
89
|
+
if (!experiment) {
|
|
90
|
+
throw new Errors.IncurError({
|
|
91
|
+
code: "INVALID_EXPERIMENT",
|
|
92
|
+
message: `Invalid experiment at ${path}.`,
|
|
93
|
+
exitCode: 1,
|
|
94
|
+
});
|
|
95
|
+
}
|
|
96
|
+
return { experimentDir: path, ...experiment };
|
|
97
|
+
}
|
|
98
|
+
const suite = await loadSuiteRun(path);
|
|
99
|
+
if (suite) {
|
|
100
|
+
return { suiteDir: path, ...suite };
|
|
101
|
+
}
|
|
102
|
+
return loadRunDetail(path);
|
|
103
|
+
}
|
|
104
|
+
|
|
105
|
+
const config = await loadLuxConfig(process.cwd());
|
|
106
|
+
const runsRoot = options.runsRoot ? resolve(options.runsRoot) : config.runsRoot;
|
|
107
|
+
const readErrors: Array<{ path: string; message: string }> = [];
|
|
108
|
+
const listingOptions = {
|
|
109
|
+
onReadError: (path: string, error: unknown) => {
|
|
110
|
+
readErrors.push({ path, message: error instanceof Error ? error.message : String(error) });
|
|
111
|
+
},
|
|
112
|
+
};
|
|
113
|
+
if (options.experiments) {
|
|
114
|
+
const experiments = await listExperimentRuns(runsRoot, listingOptions);
|
|
115
|
+
return { runsRoot, experiments, ...(readErrors.length ? { readErrors } : {}) };
|
|
116
|
+
}
|
|
117
|
+
if (options.suites) {
|
|
118
|
+
const suites = await listSuiteRuns(runsRoot, listingOptions);
|
|
119
|
+
return { runsRoot, suites, ...(readErrors.length ? { readErrors } : {}) };
|
|
120
|
+
}
|
|
121
|
+
const all = await listRuns(runsRoot);
|
|
122
|
+
const runs = options.case ? all.filter((run) => run.caseSlug === options.case) : all;
|
|
123
|
+
return {
|
|
124
|
+
runsRoot,
|
|
125
|
+
...(!options.case ? { suites: await listSuiteRuns(runsRoot, listingOptions) } : {}),
|
|
126
|
+
runs,
|
|
127
|
+
...(readErrors.length ? { readErrors } : {}),
|
|
128
|
+
};
|
|
129
|
+
};
|
|
@@ -0,0 +1,211 @@
|
|
|
1
|
+
import { realpath } from "node:fs/promises";
|
|
2
|
+
import { resolve, sep } from "node:path";
|
|
3
|
+
import type { AgentDriver, Artifact, DriverEventMap, StartContext } from "../core/driver.js";
|
|
4
|
+
import type { ContainerExecution } from "../core/driver-execution.js";
|
|
5
|
+
import { TypedEmitter } from "../core/emitter.js";
|
|
6
|
+
import {
|
|
7
|
+
type Answer,
|
|
8
|
+
DoneSchema,
|
|
9
|
+
ErrorEventSchema,
|
|
10
|
+
InterruptSchema,
|
|
11
|
+
ProgressSchema,
|
|
12
|
+
type ReadyState,
|
|
13
|
+
ReadyStateSchema,
|
|
14
|
+
} from "../core/events.js";
|
|
15
|
+
import { type ManagedProcess, spawnManagedProcess } from "../core/exec.js";
|
|
16
|
+
import { walkArtifacts } from "../drivers/shared/artifacts.js";
|
|
17
|
+
import { LUX_VERSION } from "../orchestrator/package-version.js";
|
|
18
|
+
import { type Message, readMessages, writeMessage } from "./protocol.js";
|
|
19
|
+
import { prepareContainer } from "./resources.js";
|
|
20
|
+
|
|
21
|
+
export class ContainerDriver extends TypedEmitter<DriverEventMap> implements AgentDriver {
|
|
22
|
+
private process: ManagedProcess | undefined;
|
|
23
|
+
private resources: Awaited<ReturnType<typeof prepareContainer>> | undefined;
|
|
24
|
+
private sequence = 0;
|
|
25
|
+
private pending = new Map<
|
|
26
|
+
number,
|
|
27
|
+
{ resolve: (value: unknown) => void; reject: (error: Error) => void }
|
|
28
|
+
>();
|
|
29
|
+
private closing: Promise<void> | undefined;
|
|
30
|
+
private context: StartContext | undefined;
|
|
31
|
+
private terminal = false;
|
|
32
|
+
private readonly abort = () => {
|
|
33
|
+
void this.close().catch(() => {});
|
|
34
|
+
};
|
|
35
|
+
|
|
36
|
+
private readonly name: string;
|
|
37
|
+
private readonly config: unknown;
|
|
38
|
+
private readonly execution: ContainerExecution;
|
|
39
|
+
constructor(name: string, config: unknown, execution: ContainerExecution) {
|
|
40
|
+
super();
|
|
41
|
+
this.name = name;
|
|
42
|
+
this.config = config;
|
|
43
|
+
this.execution = execution;
|
|
44
|
+
JSON.stringify(config, (_key, value: unknown) => {
|
|
45
|
+
if (typeof value === "function" || typeof value === "symbol" || typeof value === "bigint")
|
|
46
|
+
throw new Error("container driver configuration must be JSON serializable");
|
|
47
|
+
return value;
|
|
48
|
+
});
|
|
49
|
+
}
|
|
50
|
+
|
|
51
|
+
private fail = (error: Error): void => {
|
|
52
|
+
for (const pending of this.pending.values()) pending.reject(error);
|
|
53
|
+
this.pending.clear();
|
|
54
|
+
if (!this.terminal && !this.closing) {
|
|
55
|
+
this.terminal = true;
|
|
56
|
+
void this.emit("error", { exitReason: "error", reason: error.message }).catch(() => {});
|
|
57
|
+
}
|
|
58
|
+
};
|
|
59
|
+
|
|
60
|
+
private receive = (message: Message): void => {
|
|
61
|
+
if (message.event) {
|
|
62
|
+
const dispatch = async () => {
|
|
63
|
+
switch (message.event) {
|
|
64
|
+
case "interrupt":
|
|
65
|
+
await this.emit("interrupt", InterruptSchema.parse(message.value));
|
|
66
|
+
break;
|
|
67
|
+
case "progress":
|
|
68
|
+
await this.emit("progress", ProgressSchema.parse(message.value));
|
|
69
|
+
break;
|
|
70
|
+
case "done":
|
|
71
|
+
this.terminal = true;
|
|
72
|
+
await this.emit("done", DoneSchema.parse(message.value));
|
|
73
|
+
break;
|
|
74
|
+
case "error":
|
|
75
|
+
this.terminal = true;
|
|
76
|
+
await this.emit("error", ErrorEventSchema.parse(message.value));
|
|
77
|
+
break;
|
|
78
|
+
default:
|
|
79
|
+
throw new Error("unknown container event");
|
|
80
|
+
}
|
|
81
|
+
if (this.process) writeMessage(this.process.child.stdin, { id: message.id, ack: true });
|
|
82
|
+
};
|
|
83
|
+
void dispatch().catch(this.fail);
|
|
84
|
+
return;
|
|
85
|
+
}
|
|
86
|
+
const pending = this.pending.get(message.id);
|
|
87
|
+
if (!pending) return;
|
|
88
|
+
this.pending.delete(message.id);
|
|
89
|
+
if (message.error !== undefined) pending.reject(new Error(message.error));
|
|
90
|
+
else pending.resolve(message.value);
|
|
91
|
+
};
|
|
92
|
+
|
|
93
|
+
private rpc = (method: string, value?: unknown, timeoutMs = 30_000): Promise<unknown> =>
|
|
94
|
+
new Promise((resolvePromise, reject) => {
|
|
95
|
+
if (!this.process) {
|
|
96
|
+
reject(new Error("container process is unavailable"));
|
|
97
|
+
return;
|
|
98
|
+
}
|
|
99
|
+
const id = ++this.sequence;
|
|
100
|
+
const timer = setTimeout(() => {
|
|
101
|
+
this.pending.delete(id);
|
|
102
|
+
reject(new Error(`container ${method} timed out`));
|
|
103
|
+
}, timeoutMs);
|
|
104
|
+
this.pending.set(id, {
|
|
105
|
+
resolve: (result) => {
|
|
106
|
+
clearTimeout(timer);
|
|
107
|
+
resolvePromise(result);
|
|
108
|
+
},
|
|
109
|
+
reject: (error) => {
|
|
110
|
+
clearTimeout(timer);
|
|
111
|
+
reject(error);
|
|
112
|
+
},
|
|
113
|
+
});
|
|
114
|
+
try {
|
|
115
|
+
writeMessage(this.process.child.stdin, { id, method, value });
|
|
116
|
+
} catch (error) {
|
|
117
|
+
this.pending.get(id)?.reject(error as Error);
|
|
118
|
+
this.pending.delete(id);
|
|
119
|
+
}
|
|
120
|
+
});
|
|
121
|
+
|
|
122
|
+
async start(ctx: StartContext): Promise<ReadyState> {
|
|
123
|
+
this.context = ctx;
|
|
124
|
+
if (ctx.abortSignal.aborted) throw new Error("container run aborted before startup");
|
|
125
|
+
const raw = this.config as { cwd?: string; cwdMode?: string };
|
|
126
|
+
if (raw?.cwdMode === "runDir")
|
|
127
|
+
throw new Error("container drivers must use the artifacts workspace, not cwdMode: runDir");
|
|
128
|
+
if (raw?.cwd && (await realpath(raw.cwd)) !== (await realpath(ctx.artifactsDir)))
|
|
129
|
+
throw new Error("container driver cwd must be the mounted artifacts workspace");
|
|
130
|
+
this.resources = await prepareContainer(this.execution, ctx.artifactsDir);
|
|
131
|
+
try {
|
|
132
|
+
this.process = spawnManagedProcess(this.resources.command, this.resources.args, {
|
|
133
|
+
env: this.resources.env,
|
|
134
|
+
});
|
|
135
|
+
readMessages(this.process.child.stdout, this.receive, this.fail);
|
|
136
|
+
// Harness stderr can contain secrets. The normalized driver protocol is the evidence surface.
|
|
137
|
+
this.process.child.stderr.resume();
|
|
138
|
+
void this.process.exited.then(() => this.fail(new Error("container worker exited")));
|
|
139
|
+
ctx.abortSignal.addEventListener("abort", this.abort, { once: true });
|
|
140
|
+
await this.process.spawned;
|
|
141
|
+
if (ctx.abortSignal.aborted) throw new Error("container run aborted during startup");
|
|
142
|
+
const { abortSignal: _signal, ...context } = ctx;
|
|
143
|
+
const ready = ReadyStateSchema.parse(
|
|
144
|
+
await this.rpc(
|
|
145
|
+
"start",
|
|
146
|
+
{
|
|
147
|
+
version: LUX_VERSION,
|
|
148
|
+
name: this.name,
|
|
149
|
+
config: this.config,
|
|
150
|
+
module: this.execution.driverModule,
|
|
151
|
+
context: {
|
|
152
|
+
...context,
|
|
153
|
+
artifactsDir: this.resources.root,
|
|
154
|
+
...(context.fixtures
|
|
155
|
+
? { fixtures: { ...context.fixtures, rootDir: this.resources.root } }
|
|
156
|
+
: {}),
|
|
157
|
+
runDir: "/tmp/lux-run",
|
|
158
|
+
runsRoot: "/tmp/lux-runs",
|
|
159
|
+
},
|
|
160
|
+
},
|
|
161
|
+
this.execution.startupTimeoutMs ?? 60_000,
|
|
162
|
+
),
|
|
163
|
+
);
|
|
164
|
+
return {
|
|
165
|
+
...ready,
|
|
166
|
+
detail: { ...ready.detail, execution: "container", image: this.resources.image },
|
|
167
|
+
};
|
|
168
|
+
} catch (error) {
|
|
169
|
+
await this.close();
|
|
170
|
+
throw error;
|
|
171
|
+
}
|
|
172
|
+
}
|
|
173
|
+
|
|
174
|
+
async respond(answer: Answer): Promise<void> {
|
|
175
|
+
await this.rpc("respond", answer);
|
|
176
|
+
}
|
|
177
|
+
async quiesce(): Promise<void> {
|
|
178
|
+
await this.rpc("quiesce");
|
|
179
|
+
}
|
|
180
|
+
async collect(): Promise<Artifact[]> {
|
|
181
|
+
const root = this.resources?.root;
|
|
182
|
+
if (!root) return [];
|
|
183
|
+
try {
|
|
184
|
+
const artifacts = (await this.rpc("collect")) as Artifact[];
|
|
185
|
+
for (const artifact of artifacts) {
|
|
186
|
+
const path = resolve(artifact.root ?? root, artifact.path);
|
|
187
|
+
if (path !== root && !path.startsWith(`${root}${sep}`))
|
|
188
|
+
throw new Error("container driver returned artifacts outside its mounted workspace");
|
|
189
|
+
}
|
|
190
|
+
} finally {
|
|
191
|
+
await this.close();
|
|
192
|
+
}
|
|
193
|
+
return walkArtifacts(root);
|
|
194
|
+
}
|
|
195
|
+
close(): Promise<void> {
|
|
196
|
+
this.closing ??= (async () => {
|
|
197
|
+
this.context?.abortSignal.removeEventListener("abort", this.abort);
|
|
198
|
+
try {
|
|
199
|
+
if (this.process) {
|
|
200
|
+
await this.rpc("close", undefined, 5_000).catch(() => {});
|
|
201
|
+
this.process.child.stdin.end();
|
|
202
|
+
}
|
|
203
|
+
await this.resources?.cleanup();
|
|
204
|
+
} finally {
|
|
205
|
+
await this.process?.terminate();
|
|
206
|
+
this.fail(new Error("container driver closed"));
|
|
207
|
+
}
|
|
208
|
+
})();
|
|
209
|
+
return this.closing;
|
|
210
|
+
}
|
|
211
|
+
}
|