add-table-to-offline-profile

作者: microsoft

當使用者想要將一個資料表(通常是新新增的 Dataverse 資料表)加入現有的離線設定檔,而不需重新執行完整…

npx skills add https://github.com/microsoft/power-platform-skills --skill add-table-to-offline-profile

Shared instructions: shared-instructions.md — read first.

References:

Add Table to Offline Profile

Add a single table to an existing Mobile Offline Profile. Most common flow: user ran /add-dataverse to add a new table to their app, now wants that table available offline too.

Equivalent to running /setup-offline-profile and seeing the existing profile (extend-mode), but skips the per-table questionnaire for the tables already in the profile — only configures the new one.

Workflow

  1. Verify project + locate profile → 2. Resolve target table → 3. Prereq check (auto-enable if needed) → 4. Scope picker (single question) → 5. POST item + PATCH selectedcolumns → 6. Publish → 7. Update artifacts → 8. Summary

Step 1 — Verify project + locate profile

test -f power.config.json
node "${PLUGIN_ROOT}/scripts/resolve-environment.js" "$(node -e \"console.log(require('./power.config.json').environmentId)\")"

Manifest path (dual-location):

MANIFEST=$(test -f .datamodel-manifest.json && echo ".datamodel-manifest.json" || \
           (test -f docs/plan-artifacts/.datamodel-manifest.json && echo "docs/plan-artifacts/.datamodel-manifest.json"))
test -n "$MANIFEST" && echo "✓ manifest at $MANIFEST"

Profile ID resolution (same as /edit-offline-profile).

STOP if no profile exists. Recommend: Run /setup-offline-profile first to create the profile.

Step 2 — Resolve target table

$ARGUMENTS parsing:

FlagEffect
--table <logical-name>Explicit target
--all-newAdd ALL tables in the manifest that aren't already in the profile (bulk mode)
(no flags)Interactive: list tables in manifest NOT yet in profile via AskUserQuestion (max 4); if more than 4 candidates, prompt user to specify by name in next message

For bulk --all-new mode, loop through Steps 3-6 for each missing table; each runs sequentially because Dataverse profile-item POSTs serialize.

Step 3 — Prereq check (auto-enable if needed)

Query the target table's EntityMetadata:

node "${PLUGIN_ROOT}/scripts/dataverse-request.js" <envUrl> GET \
  "EntityDefinitions(LogicalName='<table>')?\$select=IsAvailableOffline,ChangeTrackingEnabled,OwnershipType,IsCustomizable"
Flags stateAction
Both already trueSkip to Step 4
Either false AND IsCustomizableAuto-enable via update-entity-offline-flags.js (no prompt — the user already opted into "offline this table" by invoking this skill)
IsCustomizable.Value = falseSTOP with BLOCKED: <table> is system-managed and cannot be flagged for offline. Use a different table or accept this row is read-only-offline.

After auto-enable, single POST PublishAllXml.

Step 4 — Scope picker (single question)

Pull from manifest the target table's lookups[] to inform defaults. Run a quick row-count probe to inform reference-data classification.

AskUserQuestion with 4 options reflecting the architect's priority cascade:

Option labelMaps to
Organization rows — User's rows only (Recommended for most tables)recorddistributioncriteria: 2, recordsownedbyme: true
All records (for small reference data)recorddistributioncriteria: 1
Related rows only (for pure child tables)recorddistributioncriteria: 0
Custom — specify exact flags in next messagePrompt user with text

Recommendation in the question body should be derived from the architect's heuristics for the table being added (use the same priority cascade from offline-profile-architect.md Step 4 inline). The user can override.

Step 5 — POST item + associations + PATCH selectedcolumns

Step 5a — POST profile item

node "${PLUGIN_ROOT}/scripts/dataverse-request.js" <envUrl> POST \
  "mobileofflineprofileitems" \
  --body '{
    "name": "<DisplayName>",
    "regardingobjectid@odata.bind": "/mobileofflineprofiles(<profileId>)",
    "selectedentitytypecode": "<table>",
    "recorddistributioncriteria": <0|1|2>,
    "recordsownedbyme": <bool>,
    "recordsownedbymyteam": false,
    "recordsownedbymybusinessunit": false,
    "getrelatedentityrecords": true,
    "syncintervalinminutes": 10
  }' \
  --include-headers

Capture itemId from OData-EntityId response header.

Step 5b — POST associations for the new table's relationships

After the new item is created, walk the new table's relationships (from the manifest's lookups[] + a EntityDefinitions(LogicalName='<table>')/ManyToOneRelationships query) and POST one mobileofflineprofileitemassociation per relationship whose target is ALREADY in the profile.

Recipe per shared/references/dataverse-offline-api.md §5–§6:

node "${PLUGIN_ROOT}/scripts/dataverse-request.js" <envUrl> POST \
  "mobileofflineprofileitemassociations" \
  --body '{
    "name": "<relationshipSchemaName>",
    "relationshipdisplayname": "<relationshipSchemaName>",
    "relationshipid": "<MetadataId-from-EntityDefinitions>",
    "regardingobjectid@odata.bind": "/mobileofflineprofileitems(<newItemId>)"
  }' \
  --include-headers

Critical: do NOT include selectedrelationshipsschema in the body — server fills it in (empirical 2026-05-24).

For new tables that are recorddistributioncriteria: 0 (Related rows only), at least ONE inbound relationship MUST be included or the table will sync zero rows. The skill should validate this and warn if no associations are being created for a Related-only-scoped table.

Step 5c — PATCH selectedcolumns

Build selectedcolumns using the deterministic union from offline-profile-architect.md Step 6 (always-include ∪ lookups ∪ screen-grep'd, dedupe, sort).

node "${PLUGIN_ROOT}/scripts/dataverse-request.js" <envUrl> PATCH \
  "mobileofflineprofileitems(<itemId>)" \
  --body '{"selectedcolumns":"{\"Columns\":[...]}"}'

Step 6 — Publish (targeted PublishXml)

node "${PLUGIN_ROOT}/scripts/dataverse-request.js" <envUrl> POST \
  "PublishXml" --body '{
    "ParameterXml": "<publish><mobileofflineprofiles><mobileofflineprofile>'"$PROFILE_ID"'</mobileofflineprofile></mobileofflineprofiles></publish>"
  }'

Publishes only this profile, not the entire org's customizations. See shared/references/dataverse-offline-api.md §9 for 0x80071141 circular-relationship handling + PublishAllXml fallback.

Step 7 — Update artifacts

Append the new table's entry to offline-profile.json tables[]. The entry MUST include a schemaColumns array — the added table's full set of schema column logical names from .datamodel-manifest.json at this moment. This is the schema-reconciliation baseline that scripts/offline-profile-delta.js diffs future manifest changes against; omitting it makes the lifecycle delta check report this table under columnBaselineMissing until it is re-reconciled. It is distinct from selectedColumns (the curated sync set) — see offline-profile-reconciliation.md. Match the canonical entry shape in /setup-offline-profile Step 9a.

Append to memory-bank.md ## Offline profile block:

addedTables:
  - { at: 2026-05-19T..., table: <name>, itemId: <guid>, scope: <criterion> }

Step 8 — Summary

✓ Added <table> to profile.

  Profile      : <name>
  New item ID  : <guid>
  Scope        : <human-readable>
  Sync         : <minutes> min
  Columns      : <N> selected
  Published    : <ISO timestamp>

Total tables in profile now: <N>

Users who already have this profile assigned will receive the new table on their
next sign-in sync. To force-refresh, sign out and back in on the device.

Status code (final line)

  • DONE — table added, profile published, artifacts updated
  • DONE_WITH_CONCERNS: <list> — added with caveats (auto-enabled prereqs, publish timeout-then-success)
  • NEEDS_CONTEXT: <missing> — no profile to add to, or no target table
  • BLOCKED: <reason> — IsCustomizable=false on target, auth failure, POST/PATCH failure

What this skill does NOT do

  • Add N:N (ManyToMany) relationships to a profile item — uncommon, not covered by v0.1's recipe (which assumes N:1 / 1:N via the regardingobjectid lookup). If the user has an M:N use case, point them at the maker portal.
  • Re-architect existing tables in the profile — that's /edit-offline-profile.
  • Remove a table from the profile — manual DELETE /mobileofflineprofileitems(<itemId>) (no current skill).

來自 microsoft 的更多技能

oss-growth
microsoft
開源增長駭客角色
agent-framework-azure-ai-py
microsoft
使用Microsoft Agent Framework Python SDK(agent-framework-azure-ai)构建Azure AI Foundry代理。适用于使用AzureAIAgentsProvider创建持久化代理、使用托管工具(代码解释器、文件搜索、网络搜索)、集成MCP服务器、管理对话线程或实现流式响应。涵盖函数工具、结构化输出和多工具代理。
development
airunway-aks-setup
microsoft
在AKS上設定AI Runway——從裸叢集到執行模型。涵蓋叢集驗證、控制器安裝、GPU評估、供應商設定及首次部署。時機:「設定AI Runway」、「上線AKS叢集」、「安裝AI Runway」、「airunway設定」、「部署模型至AKS」、「在AKS上進行GPU推論」、「在AKS上設定KAITO」、「在AKS上執行LLM」、「在AKS上使用vLLM」、「在AKS上設定模型服務」、「AI Runway控制器」。
devops
appinsights-instrumentation
microsoft
使用Azure Application Insights檢測Web應用程式的指南。提供遙測模式、SDK設定與組態參考。適用時機:如何檢測應用程式、App Insights SDK、遙測模式、什麼是App Insights、Application Insights指南、檢測範例、APM最佳實踐。
devops
applicationinsights-web-ts
microsoft
使用Application Insights JavaScript SDK(@microsoft/applicationinsights-web)為瀏覽器/Web應用程式進行檢測。適用於真實使用者監控(RUM)——頁面檢視、點擊、AJAX/fetch依賴、例外、自訂事件,以及與後端OpenTelemetry追蹤關聯的瀏覽器端GenAI代理追蹤。涵蓋SDK載入器指令碼與npm設定、框架擴充(React、React Native、Angular)、點擊分析、遙測初始化器,以及從瀏覽器發出的代理/工具/模型span的OTel GenAI語意慣例。
devops
azure-ai-anomalydetector-java
microsoft
使用適用於 Java 的 Azure AI 異常偵測器 SDK 建置異常偵測應用程式。在實作單變量/多變量異常偵測、時間序列分析或 AI 驅動監控時使用。
development
azure-ai-language-conversations-py
microsoft
使用 azure-ai-language-conversations Python SDK 實作對話語言理解(CLU)。當使用 ConversationAnalysisClient 分析對話意圖與實體、建置 NLP 功能,或將語言理解整合至應用程式時使用。
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 for Python。用於機器學習工作區、作業、模型、資料集、計算資源與管線。 觸發詞:「azure-ai-ml」、「MLClient」、「workspace」、「model registry」、「training jobs」、「datasets」。
development