juzi/graphify-out/GRAPH_REPORT.md

Graph Report - juzi (2026-10-11)

Corpus Check

• 59 files · ~4,180,488 words
• Verdict: corpus is large enough that graph structure adds value.

Summary

• 1133 nodes · 2496 edges · 75 communities (57 shown, 18 thin omitted)
• Extraction: 93% EXTRACTED · 7% INFERRED · 0% AMBIGUOUS · INFERRED: 178 edges (avg confidence: 0.53)
• Token cost: 0 input · 0 output

Graph Freshness

• Built from commit: ee744b6b
• Run git rev-parse HEAD and compare to check if the graph is stale.
• Run graphify update . after code changes (no API cost).

Community Hubs (Navigation)

• face-api.js
• t
• app.js
• skin-score.js
• n
• normalize
• drawFaceExpressions
• Box
• admin.js
• __awaiter
• AgeGenderNet
• lib/accounts.js
• server.js
• files
• .predictAgeAndGender
• .runAndExtendWithFaceDetection
• 巨子测肤 · 1'me 在线肌肤测试 — UI 设计规范
• extractFaces
• 1'me 在线肌肤测试 H5
• head-pose.js
• build_fixtures.py
• disposeUnusedWeightTensors
• .loadFromDisk
• facepp-skinanalyze（旷视皮肤分析专业版调用技能）
• 巨子测肤 · 1'me 语音播报内容设计文档 V1.0
• mobile-select.iife.min.js
• wi
• gen_voice.py
• i
• face-api.js 本地化依赖（h5/vendor/faceapi/）
• ui-20261009/manifest.json
• .toString
• FaceMatcher
• build-fixture.py
• DetectSingleFaceLandmarksTask
• C
• UI 外发版 10.09 图片资源
• TinyYolov2
• UI交互界面调整
• ComputeAllFaceDescriptorsTask
• .draw
• export_psd_assets.py
• FaceLandmark68Net
• start-local.sh
• PredictAllAgeAndGenderWithFaceAlignmentTask
• PredictAllFaceExpressionsWithFaceAlignmentTask
• k
• PredictSingleFaceExpressionsWithFaceAlignmentTask
• mobile-select/README.md
• call_skinanalyze_3view.sh
• 护理建议（全量文本，按需求摘抄）
• UI 对齐 10.09 设计稿：调研与实施计划
• FaceLandmarks68
• .forwardInput
• regression.js
• .dispose
• call_skinanalyze.sh
• NeuralNetwork
• 提交前 Review（2026-10-11）
• e2e-browser.mjs
• account-client.js
• PredictSingleFaceExpressionsTask
• .runAndExtendWithFaceDetections
• ui-20261009.mjs
• ComputeSingleFaceDescriptorTask
• misans/README.md
• extractFaceTensors
• build_report_copy.py

God Nodes (most connected - your core abstractions)

1. __awaiter() - 59 edges
2. __generator() - 59 edges
3. n() - 37 edges
4. r() - 35 edges
5. t() - 33 edges
6. i() - 33 edges
7. C() - 32 edges
8. toNetInput() - 27 edges
9. e() - 22 edges
10. f() - 20 edges

Surprising Connections (you probably didn't know these)

• loadRecords() --indirect_call--> it() [INFERRED] h5/js/admin.js → h5/vendor/faceapi/face-api.js
• bind() --indirect_call--> a() [INFERRED] h5/js/admin.js → h5/vendor/faceapi/face-api.js
• detectFrame() --indirect_call--> d() [INFERRED] h5/js/app.js → h5/vendor/faceapi/face-api.js
• vizCompose() --indirect_call--> rh() [INFERRED] h5/js/app.js → h5/vendor/faceapi/face-api.js
• vizCompose() --indirect_call--> w() [INFERRED] h5/js/app.js → h5/vendor/faceapi/face-api.js

Import Cycles

• None detected.

Communities (75 total, 18 thin omitted)

Community 0 - "face-api.js"

Cohesion: 0.02 Nodes (32): cn(), computeReshapedDimensions(), createBrowserEnv(), createFileSystem(), createNodejsEnv(), dr(), en(), extractBoundingBoxes() (+24 more)

Community 1 - "t"

Cohesion: 0.08 Nodes (33): be(), bi(), Bo(), ce(), de(), e(), Ef(), er() (+25 more)

Community 2 - "app.js"

Cohesion: 0.06 Nodes (74): applyMirror(), armNudge(), beginDetect(), bindBackdropClose(), busyWithFlow(), captureFromVideo(), checkStill(), closeOverlay() (+66 more)

Community 3 - "skin-score.js"

Cohesion: 0.12 Nodes (44): acneGrade(), buildViewModel(), calcAcne(), calcContour(), calcSens(), calcSmooth(), calcTone(), calcWrinkles() (+36 more)

Community 4 - "n"

Cohesion: 0.17 Nodes (32): a(), ae(), bf(), cf(), d(), drawContour(), dt(), ea() (+24 more)

Community 5 - "normalize"

Cohesion: 0.09 Nodes (26): conv(), conv$1(), convDown(), convLayer(), convLayer$1(), convNoRelu(), convWithBatchNorm(), denseBlock3() (+18 more)

Community 6 - "drawFaceExpressions"

Cohesion: 0.22 Nodes (6): DrawBox(), drawDetections(), drawFaceExpressions(), DrawFaceLandmarks(), FaceExpressions(), isWithFaceExpressions()

Community 7 - "Box"

Cohesion: 0.13 Nodes (6): Box(), DetectAllFaceLandmarksTask(), extendWithFaceLandmarks(), FaceLandmarks(), minBbox(), ObjectDetection()

Community 8 - "admin.js"

Cohesion: 0.25 Nodes (23): api(), bind(), closeDetail(), collectFilters(), dimsHtml(), doRetry(), dSec(), el() (+15 more)

Community 9 - "__awaiter"

Cohesion: 0.14 Nodes (12): __awaiter(), extendWithFaceDescriptor(), extractSingleFaceAndComputeResult(), FaceRecognitionNet(), fetchImage(), fetchJson(), fetchNetWeights(), fetchOrThrow() (+4 more)

Community 10 - "AgeGenderNet"

Cohesion: 0.10 Nodes (6): AgeGenderNet(), extractParams$1(), FaceExpressionNet(), FaceLandmark68NetBase(), FaceLandmark68TinyNet(), FaceProcessor()

Community 11 - "lib/accounts.js"

Cohesion: 0.07 Nodes (40): build(), count(), fill(), numeric(), atomic(), body(), cookie(), cookies() (+32 more)

Community 12 - "server.js"

Cohesion: 0.08 Nodes (48): accounts, appendRecord(), callFacepp(), CRED, csvCell(), DATA_DIR, dumpRaw(), ERR_MSG (+40 more)

Community 13 - "files"

Cohesion: 0.12 Nodes (15): files, analyze-fail, analyzing, countdown-321, hint-far, hint-front, hint-hold, hint-near (+7 more)

Community 14 - ".predictAgeAndGender"

Cohesion: 0.28 Nodes (4): extendWithAge(), extendWithGender(), PredictAllAgeAndGenderTask(), PredictSingleAgeAndGenderTask()

Community 15 - ".runAndExtendWithFaceDetection"

Cohesion: 0.33 Nodes (5): allFacesMtcnn(), allFacesSsdMobilenetv1(), allFacesTinyYolov2(), detectAllFaces(), DetectSingleFaceTask()

Community 16 - "巨子测肤 · 1'me 在线肌肤测试 — UI 设计规范"

Cohesion: 0.15 Nodes (12): 1. 基准与尺寸换算, 2. 色彩规范, 3. 字体规范, 4. 组件切图(assets/svg/), 5. 图片素材(assets/images/,共 15 张), 6. 页面清单(design/ 下有整屏 PNG 预览), 7. 目录结构, 8. 还原建议 (+4 more)

Community 17 - "extractFaces"

Cohesion: 0.19 Nodes (19): awaitMediaLoaded(), bufferToImage(), createCanvas(), createCanvasFromMedia(), extractFaces(), extractImagePatches(), getContext2dOrThrow(), getEnv() (+11 more)

Community 18 - "1'me 在线肌肤测试 H5"

Cohesion: 0.07 Nodes (27): 10.09 新版 H5 实施与验收, 按确认项实现, 文档未定义部分的实施口径, 素材再生成, 验证命令, 公网服务器配置, 公网真机验收清单, 反向代理与后台保护 (+19 more)

Community 19 - "head-pose.js"

Cohesion: 0.33 Nodes (8): assertPoints(), check(), clamp(), dist(), estimate(), makeFrontal(), makeSide(), midpoint()

Community 20 - "build_fixtures.py"

Cohesion: 0.40 Nodes (9): build_case1(), build_case2(), build_truth(), col(), dump(), main(), parse_value(), set_path() (+1 more)

Community 21 - "disposeUnusedWeightTensors"

Cohesion: 0.09 Nodes (24): disposeUnusedWeightTensors(), extractorsFactory$3(), extractorsFactory$5(), extractorsFactory$7(), extractorsFactory$9(), extractParamsFromWeigthMap(), extractParamsFromWeigthMap$1(), extractParamsFromWeigthMap$2() (+16 more)

Community 22 - ".loadFromDisk"

Cohesion: 0.25 Nodes (3): ComposableTask(), getModelUris(), TinyFaceDetector()

Community 23 - "facepp-skinanalyze（旷视皮肤分析专业版调用技能）"

Cohesion: 0.20 Nodes (9): 1. 凭据（统一管理）, 2. 标准调用模板（全可选参数，每次都带）, 2b. 三视图调用模板（正脸 + 双侧脸 → 下颌线）, 3.1 本轮踩过的坑（务必避复）, 3. 三维度 × 6 分类 P/S 计算（2026-09-26 生效，取代旧六维）, 4. 产品分组公式, 5. Excel 报告铁律（必须遵守）, 6. 其他 Agent 调用约定 (+1 more)

Community 24 - "巨子测肤 · 1'me 语音播报内容设计文档 V1.0"

Cohesion: 0.22 Nodes (8): 1. 音频格式约定(已定), 2. 播报清单(14 条), 3. 播放策略(前端已实现侧约定), 4. 生成验收, A. 步骤引导(进入每个拍照步骤时播,与屏幕步骤文案一致), B. 姿态提示(实时检测不通过时播,同一提示 4 秒内不重复), C. 拍摄与流程, 巨子测肤 · 1'me 语音播报内容设计文档 V1.0

Community 25 - "mobile-select.iife.min.js"

Cohesion: 0.20 Nodes (16): _arrayLikeToArray(), _arrayWithHoles(), _arrayWithoutHoles(), _classCallCheck(), _createClass(), _defineProperties(), h(), _iterableToArray() (+8 more)

Community 26 - "wi"

Cohesion: 0.32 Nodes (8): Ci(), Ei(), Fi(), Ii(), ki(), ri(), wi(), Xt()

Community 27 - "gen_voice.py"

Cohesion: 0.30 Nodes (11): ndarray, Path, gen_countdown(), main(), normalize_clip(), 倒计时单独按字生成再拼接，避免「三、二、一」整句被读成播报腔、与提示音气质漂移。 与提示音共用同一 speaker + instruct；句间留约…, 裁掉过长首尾静音，保留 ≤150ms 边沿，避免提示显迟钝。, 响度归一 −16 LUFS + 转 MP3 48k CBR mono 22050Hz。 (+3 more)

Community 28 - "i"

Cohesion: 0.08 Nodes (32): ba(), bn(), ca(), cl(), da(), dn(), ee(), et() (+24 more)

Community 29 - "face-api.js 本地化依赖（h5/vendor/faceapi/）"

Cohesion: 0.29 Nodes (6): face-api.js 本地化依赖（h5/vendor/faceapi/）, 为什么本地化, 文件清单与体积, 来源与版本, 用法, 许可证

Community 30 - "ui-20261009/manifest.json"

Cohesion: 0.33 Nodes (5): assets, document_size, layers, pages, source

Community 31 - ".toString"

Cohesion: 0.33 Nodes (5): FaceMatch(), la(), pr(), pt(), vt()

Community 33 - "build-fixture.py"

Cohesion: 0.60 Nodes (4): main(), 递归缩放 result 内全部坐标:{left,top,width,height} 矩形与 {x,y} 点,与原数据同为整数。, scale_coords(), to_data_url()

Community 36 - "C"

Cohesion: 0.08 Nodes (29): ao(), C(), cd(), co(), Do(), eo(), fo(), Ft() (+21 more)

Community 37 - "UI 外发版 10.09 图片资源"

Cohesion: 0.40 Nodes (4): UI 外发版 10.09 图片资源, 前端素材对应, 文件位置, 重新导出

Community 39 - "UI交互界面调整"

Cohesion: 0.40 Nodes (4): 1meAI测肤旷视UI问题反馈, UI交互界面调整, 交互, 设计

Community 42 - "export_psd_assets.py"

Cohesion: 0.70 Nodes (4): bounds(), layer_at(), main(), save_image()

Community 44 - "start-local.sh"

Cohesion: 0.50 Nodes (3): HOST, PORT, start-local.sh script

Community 47 - "k"

Cohesion: 0.18 Nodes (11): b(), gn(), k(), kn(), ne(), tt(), wo(), Yt() (+3 more)

Community 53 - "护理建议（全量文本，按需求摘抄）"

Cohesion: 0.10 Nodes (19): 5.1 轮廓（紧塑力）, 5.2 皱纹（紧塑力）, 5.3 肤色（透亮感）, 5.4 平滑度（透亮感）, 5.5 水分（透亮感）, 5.6 泛红（稳韧力）, 5.7 痘痘（稳韧力）, 6.1 轮廓 (+11 more)

Community 54 - "UI 对齐 10.09 设计稿：调研与实施计划"

Cohesion: 0.11 Nodes (17): 1. 结论和依据, 2. 统一比较口径, 3. 逐页差异和调整内容, 4. 报告数据如何接到新 UI, 5. 登录与“我的报告”的真实能力缺口, 6. 原稿中需要整理的内容, 7. 实施顺序与交付物, 8. 验收定义 (+9 more)

Community 55 - "FaceLandmarks68"

Cohesion: 0.19 Nodes (3): FaceLandmarks5(), FaceLandmarks68(), getCenterPoint()

Community 56 - ".forwardInput"

Cohesion: 0.22 Nodes (9): boxPredictionLayer(), decodeBoxesLayer(), depthwiseConvLayer(), getCenterCoordinatesAndSizesLayer(), getStridesForLayerIdx(), mobileNetV1(), outputLayer(), pointwiseConvLayer() (+1 more)

Community 57 - "regression.js"

Cohesion: 0.60 Nodes (5): diffCtx(), loadJSON(), main(), okNum(), okStr()

Community 58 - ".dispose"

Cohesion: 0.15 Nodes (12): getSizesForScale(), iou(), nonMaxSuppression(), nonMaxSuppression$1(), normalize$1(), rescaleAndNormalize(), scale(), SsdMobilenetv1() (+4 more)

Community 61 - "NeuralNetwork"

Cohesion: 0.05 Nodes (32): bgrToRgbTensor(), createFaceDetectionNet(), createFaceRecognitionNet(), createMtcnn(), createSsdMobilenetv1(), createTinyFaceDetector(), createTinyYolov2(), Dimensions() (+24 more)

Community 62 - "提交前 Review（2026-10-11）"

Cohesion: 0.40 Nodes (4): 仍需公网/业务确认, 已发现并修复, 提交前 Review（2026-10-11）, 验证

Community 63 - "e2e-browser.mjs"

Cohesion: 0.29 Nodes (4): EXPECTED, fails, pageErrors, pwRoot

Community 64 - "account-client.js"

Cohesion: 0.90 Nodes (4): browser(), init(), read(), request()

Community 71 - "extractFaceTensors"

Cohesion: 0.18 Nodes (9): extendWithFaceDetection(), extendWithFaceExpressions(), extractAllFacesAndComputeResults(), extractFaceTensors(), isTensor3D(), isWithFaceDetection(), isWithFaceLandmarks(), PredictAllFaceExpressionsTask() (+1 more)

Knowledge Gaps

• 152 isolated node(s): source, document_size, pages, assets, layers (+147 more) These have ≤1 connection - possible missing edges or undocumented components.
• 18 thin communities (<3 nodes) omitted from report — run graphify query to explore isolated nodes.

Suggested Questions

Questions this graph is uniquely positioned to answer:

• Why does k() connect k to face-api.js, t, app.js, n, lib/accounts.js, i? High betweenness centrality (0.078) - this node is a cross-community bridge.
• Why does renderReport() connect app.js to t, k? High betweenness centrality (0.058) - this node is a cross-community bridge.
• Why does rate() connect lib/accounts.js to k? High betweenness centrality (0.055) - this node is a cross-community bridge.
• Are the 24 inferred relationships involving n() (e.g. with newRecordId() and ae()) actually correct? n() has 24 INFERRED edges - model-reasoned connections that need verification.
• Are the 27 inferred relationships involving r() (e.g. with handleExport() and handleRecordsQuery()) actually correct? r() has 27 INFERRED edges - model-reasoned connections that need verification.
• Are the 7 inferred relationships involving t() (e.g. with renderReport() and e()) actually correct? t() has 7 INFERRED edges - model-reasoned connections that need verification.
• What connects source, document_size, pages to the rest of the system? 152 weakly-connected nodes found - possible documentation gaps or missing edges.