mixdog 0.9.151 → 0.9.153
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/LICENSES/browser-import-NOTICE.txt +11 -0
- package/NOTICE.md +11 -0
- package/README.md +108 -68
- package/package.json +17 -4
- package/scripts/computer-task-cost-report.mjs +144 -0
- package/scripts/run-office-live-tests.mjs +32 -0
- package/scripts/tool-stress.mjs +0 -2
- package/src/agents/front-worker/AGENT.md +13 -0
- package/src/agents/front-worker/agent.json +6 -0
- package/src/agents/heavy-worker/AGENT.md +10 -11
- package/src/agents/heavy-worker/agent.json +1 -1
- package/src/agents/maintainer/AGENT.md +0 -2
- package/src/agents/reviewer/AGENT.md +9 -13
- package/src/agents/reviewer/agent.json +1 -1
- package/src/agents/security/AGENT.md +14 -0
- package/src/agents/security/agent.json +6 -0
- package/src/agents/worker/AGENT.md +8 -8
- package/src/agents/worker/agent.json +1 -1
- package/src/agents/writer/AGENT.md +14 -0
- package/src/agents/writer/agent.json +6 -0
- package/src/defaults/skills/setup/SKILL.md +18 -27
- package/src/headless-exec.mjs +3 -10
- package/src/headless-exec.test.mjs +2 -7
- package/src/rules/agent/00-common.md +4 -7
- package/src/rules/lead/01-general.md +0 -3
- package/src/rules/shared/10-tool-workflow.md +10 -2
- package/src/runtime/agent/orchestrator/agent-runtime/agent-dispatch.mjs +183 -26
- package/src/runtime/agent/orchestrator/agent-runtime/agent-dispatch.test.mjs +378 -0
- package/src/runtime/agent/orchestrator/agent-runtime/session-builder.mjs +4 -0
- package/src/runtime/agent/orchestrator/cache-break-trace.mjs +109 -0
- package/src/runtime/agent/orchestrator/cache-break-trace.test.mjs +115 -0
- package/src/runtime/agent/orchestrator/config.mjs +13 -68
- package/src/runtime/agent/orchestrator/context/collect-skills.test.mjs +153 -0
- package/src/runtime/agent/orchestrator/context/collect.mjs +49 -45
- package/src/runtime/agent/orchestrator/mcp/client.mjs +242 -15
- package/src/runtime/agent/orchestrator/mcp/features.test.mjs +66 -0
- package/src/runtime/agent/orchestrator/mcp/security.test.mjs +36 -20
- package/src/runtime/agent/orchestrator/providers/admission-scheduler-metrics.test.mjs +47 -0
- package/src/runtime/agent/orchestrator/providers/admission-scheduler.mjs +16 -3
- package/src/runtime/agent/orchestrator/providers/anthropic-sse.mjs +54 -33
- package/src/runtime/agent/orchestrator/providers/anthropic.mjs +1 -0
- package/src/runtime/agent/orchestrator/providers/antigravity-oauth.mjs +7 -0
- package/src/runtime/agent/orchestrator/providers/gemini-schema.mjs +7 -0
- package/src/runtime/agent/orchestrator/providers/gemini-stream.mjs +3 -8
- package/src/runtime/agent/orchestrator/providers/gemini.mjs +7 -0
- package/src/runtime/agent/orchestrator/providers/lib/anthropic-request-utils.mjs +19 -3
- package/src/runtime/agent/orchestrator/providers/lib/provider-replay.mjs +50 -0
- package/src/runtime/agent/orchestrator/providers/lib/sse-framing.mjs +93 -0
- package/src/runtime/agent/orchestrator/providers/media-normalization.mjs +30 -9
- package/src/runtime/agent/orchestrator/providers/media-parity.test.mjs +59 -0
- package/src/runtime/agent/orchestrator/providers/model-list-sanitize.mjs +3 -3
- package/src/runtime/agent/orchestrator/providers/openai-codex-identity.test.mjs +71 -0
- package/src/runtime/agent/orchestrator/providers/openai-codex-metadata.mjs +18 -26
- package/src/runtime/agent/orchestrator/providers/openai-compat-presets.mjs +8 -0
- package/src/runtime/agent/orchestrator/providers/openai-compat-stream.mjs +7 -0
- package/src/runtime/agent/orchestrator/providers/openai-compat-wire.mjs +25 -2
- package/src/runtime/agent/orchestrator/providers/openai-compat.mjs +35 -4
- package/src/runtime/agent/orchestrator/providers/openai-oauth-http-sse.mjs +47 -2
- package/src/runtime/agent/orchestrator/providers/openai-oauth-ws.mjs +103 -12
- package/src/runtime/agent/orchestrator/providers/openai-responses-payload.mjs +13 -0
- package/src/runtime/agent/orchestrator/providers/openai-turn-state.mjs +68 -0
- package/src/runtime/agent/orchestrator/providers/openai-ws-delta-parity.test.mjs +122 -6
- package/src/runtime/agent/orchestrator/providers/openai-ws-delta.mjs +36 -11
- package/src/runtime/agent/orchestrator/providers/openai-ws-pool.mjs +144 -93
- package/src/runtime/agent/orchestrator/providers/openai-ws-stream.mjs +18 -13
- package/src/runtime/agent/orchestrator/providers/provider-replay.test.mjs +288 -0
- package/src/runtime/agent/orchestrator/providers/sse-framing.test.mjs +719 -0
- package/src/runtime/agent/orchestrator/providers/stream-json-pool.mjs +399 -39
- package/src/runtime/agent/orchestrator/providers/stream-json-worker.mjs +26 -0
- package/src/runtime/agent/orchestrator/session/agent-loop.mjs +86 -7
- package/src/runtime/agent/orchestrator/session/agent-loop.test.mjs +222 -0
- package/src/runtime/agent/orchestrator/session/browser-snapshot-supersession.mjs +61 -0
- package/src/runtime/agent/orchestrator/session/browser-snapshot-supersession.test.mjs +83 -0
- package/src/runtime/agent/orchestrator/session/cache/read-cache.mjs +13 -0
- package/src/runtime/agent/orchestrator/session/compact/runner.mjs +26 -2
- package/src/runtime/agent/orchestrator/session/compact-policy.test.mjs +112 -0
- package/src/runtime/agent/orchestrator/session/compaction-read-reset.test.mjs +192 -0
- package/src/runtime/agent/orchestrator/session/context-usage-restart.test.mjs +72 -0
- package/src/runtime/agent/orchestrator/session/context-utils.mjs +100 -15
- package/src/runtime/agent/orchestrator/session/eager-dispatch.mjs +17 -1
- package/src/runtime/agent/orchestrator/session/eager-dispatch.test.mjs +74 -0
- package/src/runtime/agent/orchestrator/session/image-strip-recovery.mjs +12 -8
- package/src/runtime/agent/orchestrator/session/image-strip-recovery.test.mjs +20 -1
- package/src/runtime/agent/orchestrator/session/loop/compact-policy.mjs +110 -2
- package/src/runtime/agent/orchestrator/session/loop/pre-dispatch-deny.mjs +3 -16
- package/src/runtime/agent/orchestrator/session/loop/pre-dispatch-deny.test.mjs +28 -0
- package/src/runtime/agent/orchestrator/session/loop/recall-fasttrack.mjs +2 -2
- package/src/runtime/agent/orchestrator/session/loop/steering.mjs +6 -0
- package/src/runtime/agent/orchestrator/session/loop/tool-exec.mjs +15 -1
- package/src/runtime/agent/orchestrator/session/loop/tool-helpers.mjs +6 -0
- package/src/runtime/agent/orchestrator/session/manager/ask-session.mjs +111 -39
- package/src/runtime/agent/orchestrator/session/manager/ask-session.test.mjs +46 -0
- package/src/runtime/agent/orchestrator/session/manager/compaction-runner.mjs +30 -6
- package/src/runtime/agent/orchestrator/session/manager/message-sanitize.mjs +3 -0
- package/src/runtime/agent/orchestrator/session/manager/session-crud.mjs +27 -3
- package/src/runtime/agent/orchestrator/session/manager/session-lifecycle.mjs +26 -54
- package/src/runtime/agent/orchestrator/session/manager/session-lifecycle.test.mjs +63 -1
- package/src/runtime/agent/orchestrator/session/manager/status-telemetry.mjs +3 -0
- package/src/runtime/agent/orchestrator/session/manager/tool-resolution.mjs +4 -110
- package/src/runtime/agent/orchestrator/session/manager/turn-checkpoint-journal.mjs +580 -0
- package/src/runtime/agent/orchestrator/session/manager/turn-checkpoint-journal.test.mjs +593 -0
- package/src/runtime/agent/orchestrator/session/manager/turn-checkpoint.mjs +119 -190
- package/src/runtime/agent/orchestrator/session/manager/turn-interruption.mjs +115 -0
- package/src/runtime/agent/orchestrator/session/pre-send-compact.mjs +43 -8
- package/src/runtime/agent/orchestrator/session/provider-prefix-guard.mjs +129 -12
- package/src/runtime/agent/orchestrator/session/send-with-recovery.mjs +9 -2
- package/src/runtime/agent/orchestrator/session/store/listing.mjs +8 -1
- package/src/runtime/agent/orchestrator/session/store/save-worker.mjs +15 -1
- package/src/runtime/agent/orchestrator/session/store-live-publication.test.mjs +83 -0
- package/src/runtime/agent/orchestrator/session/store-summary-index.mjs +36 -2
- package/src/runtime/agent/orchestrator/session/store-summary-reader.mjs +157 -30
- package/src/runtime/agent/orchestrator/session/store-summary-reader.test.mjs +64 -0
- package/src/runtime/agent/orchestrator/session/store-summary-visibility.mjs +57 -0
- package/src/runtime/agent/orchestrator/session/store-summary-visibility.test.mjs +219 -0
- package/src/runtime/agent/orchestrator/session/store.mjs +26 -2
- package/src/runtime/agent/orchestrator/session/tool-batch.mjs +22 -0
- package/src/runtime/agent/orchestrator/tools/builtin/bash-tool-cwd-env.test.mjs +14 -0
- package/src/runtime/agent/orchestrator/tools/builtin/bash-tool.mjs +19 -10
- package/src/runtime/agent/orchestrator/tools/builtin/find-search-budget.test.mjs +30 -14
- package/src/runtime/agent/orchestrator/tools/builtin/git-command-policy.mjs +15 -4
- package/src/runtime/agent/orchestrator/tools/builtin/git-command-tool.mjs +73 -15
- package/src/runtime/agent/orchestrator/tools/builtin/git-command-tool.test.mjs +76 -2
- package/src/runtime/agent/orchestrator/tools/builtin/list-tool.mjs +24 -683
- package/src/runtime/agent/orchestrator/tools/builtin/native-search-client.mjs +64 -72
- package/src/runtime/agent/orchestrator/tools/builtin/native-search-runner.mjs +2 -5
- package/src/runtime/agent/orchestrator/tools/builtin/native-search-transport.mjs +75 -0
- package/src/runtime/agent/orchestrator/tools/builtin/native-search-transport.test.mjs +24 -0
- package/src/runtime/agent/orchestrator/tools/builtin/read-special-files.mjs +6 -6
- package/src/runtime/agent/orchestrator/tools/code-graph/build.mjs +12 -0
- package/src/runtime/agent/orchestrator/tools/code-graph/constants.mjs +4 -1
- package/src/runtime/agent/orchestrator/tools/code-graph/dispatch.mjs +24 -5
- package/src/runtime/agent/orchestrator/tools/code-graph/dispatch.test.mjs +44 -0
- package/src/runtime/agent/orchestrator/tools/code-graph/graph-binary.mjs +12 -2
- package/src/runtime/agent/orchestrator/tools/code-graph/graph-model.mjs +35 -21
- package/src/runtime/agent/orchestrator/tools/code-graph/memory-cache.mjs +28 -13
- package/src/runtime/agent/orchestrator/tools/code-graph/memory-cache.test.mjs +54 -0
- package/src/runtime/agent/orchestrator/tools/graph-manifest.json +11 -11
- package/src/runtime/agent/orchestrator/tools/patch-manifest.json +11 -11
- package/src/runtime/agent/orchestrator/tools/progress-message.mjs +12 -0
- package/src/runtime/agent/orchestrator/tools/spawn-manifest.json +11 -11
- package/src/runtime/attachments/pdf-extract.mjs +3 -0
- package/src/runtime/attachments/pdfjs-runtime.mjs +13 -0
- package/src/runtime/bridge-clients.test.mjs +827 -0
- package/src/runtime/browser-bridge/action-schema.mjs +251 -0
- package/src/runtime/browser-bridge/client.mjs +152 -0
- package/src/runtime/browser-bridge/tool-defs.mjs +159 -0
- package/src/runtime/computer-bridge/action-schema.mjs +653 -0
- package/src/runtime/computer-bridge/client.mjs +231 -0
- package/src/runtime/computer-bridge/tool-defs.mjs +33 -0
- package/src/runtime/media/store.mjs +18 -8
- package/src/runtime/media/store.test.mjs +14 -0
- package/src/runtime/memory/index.mjs +1 -1
- package/src/runtime/memory/lib/transcript-ingest.mjs +8 -6
- package/src/runtime/memory/lib/transcript-ingest.test.mjs +52 -0
- package/src/runtime/office/assurance-benchmark.mjs +358 -0
- package/src/runtime/office/assurance.mjs +833 -0
- package/src/runtime/office/benchmark.mjs +496 -0
- package/src/runtime/office/capabilities.mjs +551 -0
- package/src/runtime/office/com-adapter.mjs +283 -0
- package/src/runtime/office/content-model.mjs +183 -0
- package/src/runtime/office/contract-benchmark.mjs +105 -0
- package/src/runtime/office/design-library.mjs +1562 -0
- package/src/runtime/office/design-system.mjs +1966 -0
- package/src/runtime/office/index.mjs +2413 -0
- package/src/runtime/office/journal.mjs +82 -0
- package/src/runtime/office/office-assurance.test.mjs +371 -0
- package/src/runtime/office/office-com-host.ps1 +3781 -0
- package/src/runtime/office/office-com-session-host.ps1 +855 -0
- package/src/runtime/office/office-live-runtime.test.mjs +819 -0
- package/src/runtime/office/office-runtime.test.mjs +2070 -0
- package/src/runtime/office/ooxml-validator.mjs +271 -0
- package/src/runtime/office/pagination.mjs +132 -0
- package/src/runtime/office/pdf-adapter.mjs +591 -0
- package/src/runtime/office/pdf-analysis.mjs +372 -0
- package/src/runtime/office/pdf-render.mjs +167 -0
- package/src/runtime/office/pdf-security.mjs +63 -0
- package/src/runtime/office/polish-benchmark.mjs +190 -0
- package/src/runtime/office/portable-ooxml.mjs +1867 -0
- package/src/runtime/office/quality-live-benchmark.mjs +486 -0
- package/src/runtime/office/quality-pipeline.mjs +136 -0
- package/src/runtime/office/tabular.mjs +302 -0
- package/src/runtime/office/templates/mixdog-executive.pptx +0 -0
- package/src/runtime/office/templates/mixdog-executive.pptx.mixdog.json +190 -0
- package/src/runtime/office/tool-defs.mjs +113 -0
- package/src/runtime/office/visual-diff.mjs +89 -0
- package/src/runtime/office/xlsx-assertions.mjs +172 -0
- package/src/runtime/office/xlsx-contract.mjs +147 -0
- package/src/runtime/shared/agent-route-config.mjs +9 -0
- package/src/runtime/shared/child-spawn-gate.mjs +17 -0
- package/src/runtime/shared/child-spawn-remote.mjs +69 -7
- package/src/runtime/shared/child-spawn-remote.test.mjs +124 -0
- package/src/runtime/shared/pristine-execution-contract.json +2 -1
- package/src/runtime/shared/provider-api-key.mjs +1 -0
- package/src/runtime/shared/session-runtime-health.mjs +91 -0
- package/src/runtime/shared/session-runtime-health.test.mjs +53 -0
- package/src/runtime/shared/skill-document.mjs +86 -0
- package/src/runtime/shared/skill-document.test.mjs +81 -0
- package/src/runtime/shared/tool-execution-contract.mjs +16 -0
- package/src/runtime/shared/tool-surface.mjs +27 -0
- package/src/runtime/shared/turn-snapshot-store.mjs +1 -0
- package/src/runtime/shared/turn-snapshot.mjs +32 -9
- package/src/runtime/shared/user-cwd.mjs +33 -14
- package/src/runtime/shared/user-cwd.test.mjs +43 -0
- package/src/session-runtime/agent-disable.test.mjs +20 -1
- package/src/session-runtime/context-status.mjs +9 -3
- package/src/session-runtime/cwd-plugins.mjs +26 -26
- package/src/session-runtime/cwd-plugins.test.mjs +75 -0
- package/src/session-runtime/cwd-tool-routing.test.mjs +5 -0
- package/src/session-runtime/deferred-tool-delta.mjs +104 -0
- package/src/session-runtime/deferred-tool-delta.test.mjs +40 -0
- package/src/session-runtime/global-extensions.mjs +26 -0
- package/src/session-runtime/global-extensions.test.mjs +28 -0
- package/src/session-runtime/global-settings-session-identity.test.mjs +76 -0
- package/src/session-runtime/goal-runtime.mjs +1141 -0
- package/src/session-runtime/goal-runtime.test.mjs +645 -0
- package/src/session-runtime/lifecycle-api.mjs +58 -11
- package/src/session-runtime/lifecycle-api.test.mjs +183 -0
- package/src/session-runtime/mcp-glue.mjs +49 -47
- package/src/session-runtime/plugin-mcp-standard.test.mjs +177 -0
- package/src/session-runtime/plugin-mcp.mjs +96 -22
- package/src/session-runtime/prewarm.mjs +1 -10
- package/src/session-runtime/provider-request-snapshot.mjs +33 -9
- package/src/session-runtime/provider-request-snapshot.test.mjs +27 -5
- package/src/session-runtime/resource-api-global.test.mjs +109 -0
- package/src/session-runtime/resource-api.mjs +168 -116
- package/src/session-runtime/runtime-core.mjs +208 -29
- package/src/session-runtime/runtime-tunables.mjs +6 -8
- package/src/session-runtime/session-lifecycle.mjs +34 -12
- package/src/session-runtime/session-title.mjs +2 -2
- package/src/session-runtime/session-turn-api.mjs +37 -25
- package/src/session-runtime/skills-api.mjs +79 -25
- package/src/session-runtime/skills-api.test.mjs +99 -0
- package/src/session-runtime/tool-catalog-data.mjs +1 -1
- package/src/session-runtime/tool-catalog-schema.mjs +0 -5
- package/src/session-runtime/tool-catalog.mjs +42 -112
- package/src/session-runtime/tool-defs.mjs +6 -1
- package/src/session-runtime/tool-policy-surface.test.mjs +93 -17
- package/src/session-runtime/tool-surface.mjs +30 -8
- package/src/session-runtime/workflow-agents-api.mjs +8 -25
- package/src/session-runtime/workflow.mjs +8 -6
- package/src/standalone/agent-dispatch-broker.mjs +5 -9
- package/src/standalone/agent-tool/helpers.mjs +2 -2
- package/src/standalone/agent-tool/job-views.mjs +3 -1
- package/src/standalone/agent-tool/notify.mjs +11 -9
- package/src/standalone/agent-tool/notify.test.mjs +37 -0
- package/src/standalone/agent-tool/render.mjs +3 -0
- package/src/standalone/agent-tool/spawn-flow.mjs +61 -16
- package/src/standalone/agent-tool/spawn-flow.test.mjs +41 -0
- package/src/standalone/agent-tool/tool-def.mjs +1 -1
- package/src/standalone/agent-tool/worker-index.mjs +123 -0
- package/src/standalone/agent-tool/worker-index.test.mjs +132 -0
- package/src/standalone/agent-tool/worker-rows.mjs +2 -1
- package/src/standalone/agent-tool.mjs +57 -10
- package/src/standalone/daemon.mjs +113 -15
- package/src/standalone/hook-bus/config.mjs +1 -0
- package/src/standalone/plugin-admin.mjs +17 -0
- package/src/standalone/plugin-admin.test.mjs +7 -0
- package/src/standalone/provider-admin.mjs +1 -0
- package/src/standalone/session-protocol.mjs +5 -0
- package/src/standalone/session-runtime-agent-control-client.mjs +133 -0
- package/src/standalone/session-runtime-agent-control-client.test.mjs +71 -0
- package/src/standalone/session-runtime-dispatch-cancel.test.mjs +139 -0
- package/src/standalone/session-runtime-host-factory.mjs +17 -0
- package/src/standalone/session-runtime-host-factory.test.mjs +31 -0
- package/src/standalone/session-runtime-host-health.test.mjs +5 -0
- package/src/standalone/session-runtime-host.mjs +709 -58
- package/src/standalone/session-runtime-inline-host.mjs +258 -0
- package/src/standalone/session-runtime-inline-host.test.mjs +385 -0
- package/src/standalone/session-runtime-provider-cooldown.mjs +78 -0
- package/src/standalone/session-runtime-provider-cooldown.test.mjs +80 -0
- package/src/standalone/session-runtime-shard-host.test.mjs +642 -0
- package/src/standalone/session-runtime-shard-router.mjs +149 -0
- package/src/standalone/session-runtime-shard-router.test.mjs +101 -0
- package/src/standalone/session-runtime-worker.mjs +302 -26
- package/src/standalone/session-service-agent-tree.test.mjs +984 -0
- package/src/standalone/session-service-goal-recovery.test.mjs +177 -0
- package/src/standalone/session-service.mjs +711 -11
- package/src/tui/app/app-view.jsx +1 -0
- package/src/tui/app/core-memory-picker.mjs +2 -0
- package/src/tui/app/extension-pickers.mjs +1 -1
- package/src/tui/app/maintenance-pickers.mjs +3 -3
- package/src/tui/app/model-options.mjs +2 -0
- package/src/tui/app/model-picker.mjs +14 -4
- package/src/tui/app/panel-handoff.test.mjs +311 -0
- package/src/tui/app/project-picker.mjs +1 -0
- package/src/tui/app/route-pickers.mjs +21 -6
- package/src/tui/app/settings-picker.mjs +26 -3
- package/src/tui/app/slash-commands.mjs +1 -0
- package/src/tui/app/slash-dispatch.mjs +73 -22
- package/src/tui/app/theme-effort-pickers.mjs +7 -2
- package/src/tui/components/Picker.jsx +10 -1
- package/src/tui/dist/index.mjs +187 -45
- package/src/tui/lib/voice-setup.mjs +4 -1
- package/src/tui/session/agent-message-source.test.mjs +37 -0
- package/src/tui/session/completion-card-restore.test.mjs +50 -0
- package/src/tui/session/context-state.mjs +3 -13
- package/src/tui/session/goal-continuation.mjs +141 -0
- package/src/tui/session/goal-continuation.test.mjs +203 -0
- package/src/tui/session/queue-helpers.mjs +5 -2
- package/src/tui/session/session-action-surface.test.mjs +190 -0
- package/src/tui/session/session-api-ext.mjs +48 -8
- package/src/tui/session/session-api.mjs +73 -1
- package/src/tui/session/session-flow.mjs +31 -1
- package/src/tui/session/turn.mjs +88 -15
- package/src/tui/session-local.mjs +39 -3
- package/src/vendor/statusline/src/gateway/route-meta.mjs +31 -0
- package/src/runtime/agent/orchestrator/tools/builtin/fuzzy-match.mjs +0 -301
|
@@ -0,0 +1,190 @@
|
|
|
1
|
+
import assert from 'node:assert/strict';
|
|
2
|
+
import { mkdtemp, rm } from 'node:fs/promises';
|
|
3
|
+
import { tmpdir } from 'node:os';
|
|
4
|
+
import { join } from 'node:path';
|
|
5
|
+
import { pathToFileURL } from 'node:url';
|
|
6
|
+
|
|
7
|
+
import { executeOfficeTool, resetOfficeSessionsForTest } from './index.mjs';
|
|
8
|
+
|
|
9
|
+
const BASELINE = Object.freeze({
|
|
10
|
+
capturedAt: '2026-08-28',
|
|
11
|
+
modelFacingCalls: 22,
|
|
12
|
+
durationMsLowerBound: 22_516.8,
|
|
13
|
+
scenarios: {
|
|
14
|
+
docx: { calls: 9, retries: 2, durationMsLowerBound: 7_634.23, accurate: false },
|
|
15
|
+
xlsx: { calls: 6, retries: 1, durationMsLowerBound: 7_514.49, accurate: false },
|
|
16
|
+
pptx: { calls: 7, retries: 0, durationMsLowerBound: 7_368.08, accurate: true },
|
|
17
|
+
},
|
|
18
|
+
});
|
|
19
|
+
|
|
20
|
+
function resultValue(result) {
|
|
21
|
+
const text = result?.content?.[0]?.text || '';
|
|
22
|
+
if (result?.isError) throw new Error(text);
|
|
23
|
+
return JSON.parse(text);
|
|
24
|
+
}
|
|
25
|
+
|
|
26
|
+
function duration(value) {
|
|
27
|
+
return Number(value?.metrics?.durationMs || 0);
|
|
28
|
+
}
|
|
29
|
+
|
|
30
|
+
async function removeBenchmarkRoot(root) {
|
|
31
|
+
let lastError = null;
|
|
32
|
+
for (let attempt = 0; attempt < 20; attempt += 1) {
|
|
33
|
+
try {
|
|
34
|
+
await rm(root, { recursive: true, force: true });
|
|
35
|
+
return;
|
|
36
|
+
} catch (error) {
|
|
37
|
+
lastError = error;
|
|
38
|
+
if (error?.code !== 'EBUSY' && error?.code !== 'EPERM') throw error;
|
|
39
|
+
await new Promise((resolve) => setTimeout(resolve, 250));
|
|
40
|
+
}
|
|
41
|
+
}
|
|
42
|
+
throw lastError;
|
|
43
|
+
}
|
|
44
|
+
|
|
45
|
+
async function runScenario({
|
|
46
|
+
root,
|
|
47
|
+
name,
|
|
48
|
+
format,
|
|
49
|
+
operations,
|
|
50
|
+
pages,
|
|
51
|
+
auditProfile,
|
|
52
|
+
assertCreated,
|
|
53
|
+
assertFinalized,
|
|
54
|
+
}) {
|
|
55
|
+
const path = join(root, `optimized-${name}.${format}`);
|
|
56
|
+
const finalized = resultValue(await executeOfficeTool({
|
|
57
|
+
action: 'create',
|
|
58
|
+
path,
|
|
59
|
+
format,
|
|
60
|
+
mode: 'background',
|
|
61
|
+
overwrite: true,
|
|
62
|
+
operations,
|
|
63
|
+
finalize: true,
|
|
64
|
+
failOn: 'warning',
|
|
65
|
+
autoFix: true,
|
|
66
|
+
...(pages ? { pages } : {}),
|
|
67
|
+
...(auditProfile ? { auditProfile } : {}),
|
|
68
|
+
maxWidth: 900,
|
|
69
|
+
}, { cwd: root }));
|
|
70
|
+
assert.equal(finalized.batch.changeSummary.noChange, 0);
|
|
71
|
+
assert.equal(finalized.batch.changeSummary.changed, operations.length);
|
|
72
|
+
assertCreated(finalized);
|
|
73
|
+
assert.equal(finalized.ok, true, JSON.stringify(finalized));
|
|
74
|
+
assert.equal(finalized.finalized, true);
|
|
75
|
+
assert.equal(finalized.validation.ok, true);
|
|
76
|
+
assert.equal(finalized.validation.native.documentSaved, true);
|
|
77
|
+
assert.equal(finalized.closed, true);
|
|
78
|
+
if (assertFinalized) assertFinalized(finalized);
|
|
79
|
+
return {
|
|
80
|
+
name,
|
|
81
|
+
format,
|
|
82
|
+
calls: 1,
|
|
83
|
+
retries: 0,
|
|
84
|
+
accurate: true,
|
|
85
|
+
durationMs: duration(finalized),
|
|
86
|
+
completionDurationMs: duration(finalized),
|
|
87
|
+
issueCount: finalized.review?.issuesAfter?.length || 0,
|
|
88
|
+
finalizeSteps: finalized.stepMetrics,
|
|
89
|
+
path,
|
|
90
|
+
};
|
|
91
|
+
}
|
|
92
|
+
|
|
93
|
+
export async function runOfficePolishBenchmark({ keep = false } = {}) {
|
|
94
|
+
const root = await mkdtemp(join(tmpdir(), 'mixdog-office-polish-'));
|
|
95
|
+
try {
|
|
96
|
+
const results = [];
|
|
97
|
+
results.push(await runScenario({
|
|
98
|
+
root,
|
|
99
|
+
name: 'report',
|
|
100
|
+
format: 'docx',
|
|
101
|
+
operations: [
|
|
102
|
+
{ op: 'append_text', text: 'Office Use 운영 효율 보고서', style: 'Title' },
|
|
103
|
+
{ op: 'append_text', text: '요약', style: 'Heading 1' },
|
|
104
|
+
{ op: 'append_text', text: '처리 속도와 모델-facing Office 호출 수를 줄이면서 실제 산출물 정확도를 유지합니다.' },
|
|
105
|
+
{ op: 'append_text', text: '검증 원칙', style: 'Heading 1' },
|
|
106
|
+
{ op: 'append_text', text: '실제 Microsoft Office 재열기, 구조 검사, 시각 QA와 no-op 검사를 모두 통과해야 합니다.' },
|
|
107
|
+
{ op: 'set_header_footer', section: 1, kind: 'primary', header: true, text: 'Mixdog · Office Use' },
|
|
108
|
+
{ op: 'add_page_numbers', section: 1, includeTotal: true },
|
|
109
|
+
],
|
|
110
|
+
pages: [1],
|
|
111
|
+
assertCreated(created) {
|
|
112
|
+
assert.equal(created.batch.changeSummary.changed, 7);
|
|
113
|
+
},
|
|
114
|
+
}));
|
|
115
|
+
results.push(await runScenario({
|
|
116
|
+
root,
|
|
117
|
+
name: 'analysis',
|
|
118
|
+
format: 'xlsx',
|
|
119
|
+
operations: [
|
|
120
|
+
{ op: 'set_range', sheet: 'Sheet1', range: 'A1:D5', values: [['분기', '매출', '비용', '영업이익'], ['1분기', 120000, 85000, 35000], ['2분기', 145000, 93000, 52000], ['3분기', 138000, 91000, 47000], ['4분기', 162000, 104000, 58000]] },
|
|
121
|
+
{ op: 'set_style', sheet: 'Sheet1', range: 'A1:D1', properties: { bold: true, color: 'FFFFFF', fillColor: '1F4E78' } },
|
|
122
|
+
{ op: 'set_style', sheet: 'Sheet1', range: 'B2:D5', properties: { numberFormat: '#,##0' } },
|
|
123
|
+
{ op: 'add_table', sheet: 'Sheet1', range: 'A1:D5', name: 'QuarterlyPerformance', style: 'TableStyleMedium2' },
|
|
124
|
+
{ op: 'add_chart', sheet: 'Sheet1', range: 'A1:D5', chartType: 'column', title: '분기별 실적', left: 360, top: 20, width: 420, height: 260 },
|
|
125
|
+
{ op: 'freeze_panes', sheet: 'Sheet1', row: 2, column: 1 },
|
|
126
|
+
{ op: 'autofit_range', sheet: 'Sheet1', range: 'A:D' },
|
|
127
|
+
],
|
|
128
|
+
assertCreated(created) {
|
|
129
|
+
const chart = created.batch.results.find((entry) => entry.op === 'add_chart');
|
|
130
|
+
assert.equal(chart.series, 3);
|
|
131
|
+
assert.equal(chart.categories, 4);
|
|
132
|
+
},
|
|
133
|
+
assertFinalized(finalized) {
|
|
134
|
+
assert.equal(finalized.review.preview.pageCount, 1);
|
|
135
|
+
assert.equal(finalized.review.preview.visualCoverage.complete, true);
|
|
136
|
+
},
|
|
137
|
+
}));
|
|
138
|
+
results.push(await runScenario({
|
|
139
|
+
root,
|
|
140
|
+
name: 'deck',
|
|
141
|
+
format: 'pptx',
|
|
142
|
+
operations: [
|
|
143
|
+
{ op: 'add_slide' },
|
|
144
|
+
{ op: 'set_slide_background', slide: 1, color: 'F4F7FB' },
|
|
145
|
+
{ op: 'add_textbox', slide: 1, text: 'Office Use 성능 개선', properties: { left: 54, top: 120, width: 620, height: 80, fontName: '맑은 고딕', fontSize: 32, bold: true, color: '17365D' } },
|
|
146
|
+
{ op: 'add_textbox', slide: 1, text: '속도 · 적은 호출 · 정확한 결과', properties: { left: 58, top: 220, width: 560, height: 44, fontName: '맑은 고딕', fontSize: 18, color: '4F6478' } },
|
|
147
|
+
{ op: 'add_shape', slide: 1, shapeType: 'rectangle', properties: { left: 58, top: 285, width: 120, height: 6, fillColor: '2F75B5', lineColor: '2F75B5' } },
|
|
148
|
+
{ op: 'add_slide' },
|
|
149
|
+
{ op: 'set_slide_background', slide: 2, color: 'FFFFFF' },
|
|
150
|
+
{ op: 'add_textbox', slide: 2, text: '개선 목표', properties: { left: 48, top: 32, width: 620, height: 52, fontName: '맑은 고딕', fontSize: 26, bold: true, color: '17365D' } },
|
|
151
|
+
{ op: 'add_shape', slide: 2, shapeType: 'rounded_rectangle', paragraphs: [{ text: '처리 시간', fontName: '맑은 고딕', fontSize: 15, color: '17365D' }, { text: '−30%', fontName: '맑은 고딕', fontSize: 28, bold: true, color: '17365D' }], properties: { left: 48, top: 130, width: 190, height: 150, fillColor: 'D9EAF7', lineColor: '9DC3E6' } },
|
|
152
|
+
{ op: 'add_shape', slide: 2, shapeType: 'rounded_rectangle', paragraphs: [{ text: 'Office 호출', fontName: '맑은 고딕', fontSize: 15, color: '17365D' }, { text: '−40%', fontName: '맑은 고딕', fontSize: 28, bold: true, color: '17365D' }], properties: { left: 265, top: 130, width: 190, height: 150, fillColor: 'E2F0D9', lineColor: 'A9D18E' } },
|
|
153
|
+
{ op: 'add_shape', slide: 2, shapeType: 'rounded_rectangle', paragraphs: [{ text: 'silent no-op', fontName: '맑은 고딕', fontSize: 15, color: '7F6000' }, { text: '0건', fontName: '맑은 고딕', fontSize: 28, bold: true, color: '7F6000' }], properties: { left: 482, top: 130, width: 190, height: 150, fillColor: 'FFF2CC', lineColor: 'FFD966' } },
|
|
154
|
+
{ op: 'set_notes', slide: 2, text: 'Source: Office Use polishing benchmark' },
|
|
155
|
+
],
|
|
156
|
+
pages: [1, 2],
|
|
157
|
+
auditProfile: 'model-backed-deck',
|
|
158
|
+
assertCreated(created) {
|
|
159
|
+
assert.equal(created.batch.changeSummary.changed, 12);
|
|
160
|
+
},
|
|
161
|
+
}));
|
|
162
|
+
|
|
163
|
+
const modelFacingCalls = results.reduce((sum, entry) => sum + entry.calls, 0);
|
|
164
|
+
const durationMs = Number(results.reduce((sum, entry) => sum + entry.durationMs, 0).toFixed(2));
|
|
165
|
+
return {
|
|
166
|
+
version: 1,
|
|
167
|
+
createdAt: new Date().toISOString(),
|
|
168
|
+
baseline: BASELINE,
|
|
169
|
+
optimized: {
|
|
170
|
+
modelFacingCalls,
|
|
171
|
+
durationMs,
|
|
172
|
+
accurate: results.every((entry) => entry.accurate),
|
|
173
|
+
retries: results.reduce((sum, entry) => sum + entry.retries, 0),
|
|
174
|
+
results,
|
|
175
|
+
},
|
|
176
|
+
improvement: {
|
|
177
|
+
callReductionPercent: Number(((1 - (modelFacingCalls / BASELINE.modelFacingCalls)) * 100).toFixed(2)),
|
|
178
|
+
durationReductionPercent: Number(((1 - (durationMs / BASELINE.durationMsLowerBound)) * 100).toFixed(2)),
|
|
179
|
+
},
|
|
180
|
+
};
|
|
181
|
+
} finally {
|
|
182
|
+
resetOfficeSessionsForTest();
|
|
183
|
+
if (!keep) await removeBenchmarkRoot(root);
|
|
184
|
+
}
|
|
185
|
+
}
|
|
186
|
+
|
|
187
|
+
if (process.argv[1] && import.meta.url === pathToFileURL(process.argv[1]).href) {
|
|
188
|
+
const report = await runOfficePolishBenchmark({ keep: process.argv.includes('--keep') });
|
|
189
|
+
process.stdout.write(`${JSON.stringify(report, null, 2)}\n`);
|
|
190
|
+
}
|