[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8,"_9":10},"actionData","errors","root",{"_150":151,"_152":153},"routes/_app.x.$handle.$name",{"_94":93,"_31":43,"_102":84,"_26":103,"_13":14,"_104":-5,"_11":105,"_106":107,"_108":109,"_110":111,"_112":113,"_20":-5,"_114":-5,"_115":-5,"_116":24,"_117":118,"_119":56,"_120":-5,"_121":-5,"_122":123},"routes/_app.x.$handle.$name.branch.$branch.version.$version.diff",{"_11":12,"_13":14,"_15":16,"_17":-5,"_18":-5},"project",{"_41":42,"_31":43,"_36":44,"_45":-5,"_46":24,"_47":48,"_49":48,"_50":51,"_52":-5,"_18":53,"_54":30,"_55":56,"_57":-5,"_58":-5,"_59":60,"_61":62,"_63":-5,"_64":-5,"_65":-5,"_66":-5,"_67":68,"_69":68,"_70":71,"_72":71,"_73":71,"_74":68,"_75":68,"_76":77,"_78":79,"_80":81,"_82":24},"version",1,"changes",[19],"aiChatInfo","userId",{"_20":21,"_22":-7,"_23":24,"_25":24,"_26":27,"_28":27},"file",{"_29":30,"_31":27,"_32":33,"_34":35,"_36":37,"_38":14,"_39":40,"_23":24},"previous","deleted",false,"deletedHere","path","main.tsx","latestPath","fileId","1fbe1bde-55cb-11f1-b114-ee650bb23af1","name","fileType","http","content","import { Hono } from \"npm:hono\";\n\nconst BACKEND = \"https://vlduy123--e68fc94055b911f1a446ee650bb23af1.web.val.run\";\nconst DASHBOARD = \"https://vlduy123--8d03963255b311f1a94aee650bb23af1.web.val.run\";\nconst DOCS = \"https://vlduy123--150d8fa455bb11f19102ee650bb23af1.web.val.run\";\nconst MOCK = \"https://vlduy123--5a73622e559a11f1a2c7ee650bb23af1.web.val.run\";\n\nconst app = new Hono();\napp.get(\"/\", (c) => c.html(HTML));\n\nconst SAP_LOGO = ``;\n\nconst HIVE_LOGO = ``;\n\nconst HTML = `\n\n\n\n\nAIHive · SAP EWM × Dify · Live Demo\n\n\n\n\n\n\n\n

\n ${HIVE_LOGO}\n

AIHive · EWM Integration

\n \n Open Dashboard ↗\n

\n\n

Live · SAP EWM × Dify × Val Town

\n Reason over millions of warehouse records
\n from a single Dify chatflow.\n

\n AIHive connects a SAP S/4HANA EWM simulation to a Dify agent through a token-aware chunking pipeline.\n Ask the agent anything — when the dataset is too big for one prompt, it transparently chunks, fetches, and synthesises.\n

\n Launch Fiori Dashboard →\n Read the Docs\n

\n\n

${SAP_LOGO}

\n \n

Dify Workflow Engine

\n \n

Val Town · Hono · SQLite

\n \n

OpenUI5 Horizon Theme

\n\n

The Stack

Four small apps, one cohesive landscape.

Each piece runs on Val Town. They mirror a real SAP landscape: source system, gateway, Fiori shell, help portal.

\n\n

01 / SOURCE

ewmMockServer

Generates SAP EWM data on demand: ~11.5M virtual warehouse orders, deliveries, waves, stock positions.

\n Hono · OData\n Open ↗\n

02 / SYSTEM

ewmBackend

The \"S/4HANA\" of the stack. SQLite persistence, seed pipelines, four chunking strategies, OData reads.

\n Hono · SQLite\n Open ↗\n

03 / FIORI

ewmDashboard

OpenUI5 frontend with the Horizon theme. ToolPage shell, ShellBar, DynamicPages — exactly what S/4HANA ships.

\n OpenUI5 · Horizon\n Open ↗\n

04 / DOCS

ewmDocs

Full reference: architecture, data model, every API endpoint, Dify integration patterns, troubleshooting.

\n Static HTML\n Open ↗\n

\n\n

Dify Chatflow

Six branches. One intent classifier.

User query → classifier → matched branch → backend → synthesised answer. Branch 5 is where chunking earns its keep.

\n\n

\n \n

\n\n

Start

Classifier

LLM

HTTP

Code

Answer

\n\n

b1 · INSTANT

System Overview

\"Give me an overview\" — KPIs, row counts, recent activity.

GET /stats → GET /seed/stats → format → answer

b2 · ONE QUERY

Browse / List Records

\"List warehouse orders for PL01-WH1\" — paginated OData read.

extract → GET /data → format → answer

b3 · DERIVED DATA

Product Search

\"What products do we have?\" — distinct materials from JSON1 extraction.

GET /products → format → answer

b4 · MASTER DATA

Plants / Warehouses

\"Show me all warehouses in NL\" — master data with filter.

route → GET /<master> → format → answer

b5 · CHUNKED — the hard one

Deep Analysis over Many Records

\"Summarise all open orders\" — too big for one prompt. Plan → chunk in SQLite → fetch → concat → Opus 200K synthesis.

plan → POST /ingest → run_id → GET /chunks → concat → synth → answer

b6 · POINT LOOKUP

Specific Record

\"Show me WO-PL01-WH1-0000001\" — extract ID, scan, present.

extract ID → GET /data → detail → answer

\n\n

Live · talking to ewmBackend now

Real data, this second.

These cards call the actual backend you'd hit from Dify. Refresh the page and the numbers move.

\n\n

GET /api/stats

AI Chunking Activity

—

total chunks generated

\n — records processed · — tokens encoded\n

GET /api/seed/stats

SQLite Snapshot

—

entity rows

\n — plants · — warehouses · — seed runs\n

GET /api/products

Distinct Materials

—

unique products

Loading top categories…

\n\n

Inside b5

How chunking solves the context problem.

User asks: \"Summarise all open warehouse orders for PL01-WH1.\" 50,000 records seeded. None of them fit in a single Claude prompt naively. Here's what happens.

\n\n

Plan the ingest

Opus reads the query, picks entity=WarehouseOrders, strategy=token-budget, tokenBudget=2000, warehouseId=PL01-WH1, limit=2000.

LLM · Opus

Backend chunks in SQLite

POST /api/ingest packs records into ~2000-token chunks server-side, persists to chunks table. Returns run_id and totals.

HTTP · POST

Extract run_id

A 10-line JS Code node pulls run_id and aggregates from the JSON for downstream use.

Code · JS

Fetch chunks with payload

GET /api/chunks?run_id=…&payload=1&limit=50 — returns up to 50 chunks, each with its JSON payload.

HTTP · GET

Concatenate & cap

Code node merges chunks with --- CHUNK n/m --- headers, caps at 30 chunks / 120K chars to stay inside Opus's 200K context.

Code · JS

Synthesise

Opus 4.7 reads the full stitched context and produces the final answer with chunk-level citations.

LLM · Opus

Reply with breadcrumb

Answer node appends \"Analysed N records across M/K chunks (run abc123…)\" so you can drill back in via the dashboard.

Answer

\n\n

Ship a chunked AI agent in an afternoon.

Import the YAML into Dify. Point the HTTP nodes at the backend. Talk to your warehouse.

\n Open the Dashboard\n Read the Docs\n

\n\n\n\n\n\n`;\n\nexport default app.fetch;\n","createdAt",["D",1779446632438],"branchVersion","branchId","1f4de1ac-55cb-11f1-9a07-ee650bb23af1","id","1f4d0e8a-55cb-11f1-9a07-ee650bb23af1","ewmDemo",["D",1779446631705],"deletedAt","legacyEnvVars","privacy","public","httpPrivacy","description","AIHive demo page for SAP EWM × Dify integration","imageUrl","e8c17657-6210-4c45-bd2b-dbcf2e020293","pinnedFileId","pinPreview",true,"pinMigrated","screenshotVersion","likeCount",0,"orgId","019e4de5-f6b1-7488-a017-a3a028bc2ba8","tursoName","tursoHostname","tursoMeasuredAt","pinned","userPerm","r","valPerm","blobPerm","rw","emailPerm","sqlitePerm","projectPerm","telemetryPerm","user",{"_94":93,"_101":97},"org",{"_94":93,"_95":93,"_96":97,"_18":53,"_98":99,"_100":24},"branch",{"_13":83,"_31":84,"_41":40,"_85":42,"_86":-5,"_87":-5,"_88":-5,"_89":-5,"_90":-5,"_91":43,"_92":93},"isOwned",2,"main","projectId","forkedBranchId","forkedBranchVersion","forkedProjectPrivacy","forkedProjectId","forkedProjectOrgId","projectName","userHandle","vlduy123","handle","displayName","avatar","https://avatars.githubusercontent.com/u/208025931?v=4","tier","free","canRestrictVals","profileImageUrl","branchName","","adoptOrgId",{"_41":42,"_31":43,"_36":146,"_45":-5,"_46":24,"_47":48,"_49":48,"_50":51,"_52":-5,"_18":53,"_54":30,"_55":56,"_57":-5,"_58":-5,"_59":60,"_61":62,"_63":-5,"_64":-5,"_65":-5,"_66":-5,"_67":68,"_69":68,"_70":71,"_72":71,"_73":71,"_74":68,"_75":68,"_76":147,"_78":148,"_80":149,"_82":24},"tree",[141],"files",[141],"counts",{"_131":60,"_132":14,"_133":60,"_134":135,"_136":137},"contents",[124],"upstreamProject","upstreamBranch","emptyEnvVars","flags",[],"hasHttpFiles","grants","orgs","customDomains",{},{"_41":125,"_31":27,"_29":30,"_126":-5,"_39":40,"_38":14,"_127":24,"_36":128,"_13":60,"_32":33,"_34":35,"_23":24,"_129":-5,"_26":130},"1fbe1b3e-55cb-11f1-b114-ee650bb23af1","parentId","latest",["D",1779446632438],"label",[27],"pulls","branches","forks","envVars",{"_138":60,"_140":60},"connections",{"_138":60,"_139":60},"count","unfilledCount","emptyCount",{"_41":125,"_31":27,"_29":30,"_126":-5,"_39":40,"_38":14,"_127":24,"_36":142,"_13":60,"_32":33,"_34":35,"_23":24,"_129":-5,"_143":144,"_26":145},["D",1779446632438],"children",[],[27],["D",1779446631705],{"_94":93,"_101":97},{"_94":93,"_95":93,"_96":97,"_18":53,"_98":99,"_100":24},{"_13":83,"_31":84,"_41":40,"_85":42,"_86":-5,"_87":-5,"_88":-5,"_89":-5,"_90":-5,"_91":43,"_92":93},"ctx",{"_76":-5,"_121":183,"_184":185,"_186":187,"_188":24,"_189":24,"_190":24,"_191":56,"_192":-5,"_193":-7,"_194":-7,"_195":60,"_196":24,"_197":24,"_198":-5,"_199":24,"_200":24,"_201":24,"_202":-5},"ENV",{"_154":155,"_156":157,"_158":84,"_159":160,"_161":162,"_163":164,"_165":166,"_167":168,"_169":170,"_171":172,"_173":174,"_175":176,"_177":178,"_179":180,"_181":182},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","e87eb04e707cf64e7e532c8e8ec395f209e17f15","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

\n Reason over millions of warehouse records\n from a single Dify chatflow.\n

Four small apps, one cohesive landscape.

ewmMockServer

ewmBackend

ewmDashboard

ewmDocs

Six branches. One intent classifier.

System Overview

Browse / List Records

Product Search

Plants / Warehouses

Deep Analysis over Many Records

Specific Record

Real data, this second.

AI Chunking Activity

SQLite Snapshot

Distinct Materials

How chunking solves the context problem.

Plan the ingest

Backend chunks in SQLite

Extract run_id

Fetch chunks with payload

Concatenate & cap

Synthesise

Reply with breadcrumb

Ship a chunked AI agent in an afternoon.

\n Reason over millions of warehouse records
\n from a single Dify chatflow.\n

Diff · v1