edit-offline-profile

작성자: microsoft

사용자가 전체를 다시 실행하지 않고 기존 오프라인 프로필의 한 가지 측면(테이블의 행 범위, 열 목록, 동기화 빈도)을 변경하려는 경우 사용합니다.

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

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

References:

Edit Offline Profile

Re-run a single piece of an existing profile — change one table's row scope, update the column list, adjust sync frequency, or rename the profile. Avoids the cognitive cost of walking the full /setup-offline-profile wizard for a one-line change.

Scope: existing profile (read from <working_dir>/offline-profile.json or --profile-id); edits at the table-item granularity. To ADD a new table return to the owner for /add-table-to-offline-profile; deleting an entire profile is outside this helper's scope, not an automatic recovery.

Invocation context and root binding

This internal helper is always a child invocation. Before any project read or command, execute app-working-directory.md. Require the owner's absolute working_dir; --working-dir, when supplied, must identify that same native-canonical root. Missing, relative, or conflicting root context returns NEEDS_CONTEXT before project access; an inaccessible directory is BLOCKED. Never use the launch cwd or introduce another root resolver.

Receive MOBILE_APP_ORCHESTRATING=1, orchestrator, working_dir, phase, approved_scope, and supplied answers from the owner. Preserve the selected environment ID/URL/tenant, profile ID, table allowlist, item IDs, and exact scope/column/sync/name/description changes. Missing or conflicting scope returns NEEDS_CONTEXT, not another environment/profile/table selection. The marker is not approval: schema approval alone does not authorize profile edits, publication, deletion, or membership edits. Retain Step 3's single confirm unless supplied answers already approve that same current offline change. For --plan-only or a planning-phase handoff, return the current/proposed diff after read-only discovery and STOP before Apply, PATCH, publication, or artifact writes. A gate response cannot lift proposal-only mode.

Apply the canonical contract to every shell call and file tool, including version/auth checks, manifest/screen reads, commands copied from references, verification, recovery, and retries. Re-supply the absolute PLUGIN_ROOT and required context on each call; no prior cd, export, or shell variable persists. Use the canonical Bash guard below or its PowerShell equivalent. Replace placeholders with validated, shell-quoted invocation values; missing values return NEEDS_CONTEXT, never execute an unresolved placeholder. File tools use absolute <working_dir>/... paths; plugin reads/scripts use ${PLUGIN_ROOT}/.... Never search another app when a file is absent.

Workflow

  1. Verify project + locate profile → 2. Resolve target (which table / what to change) → 3. Show current vs proposed → Single confirm → 4. PATCH → 5. Publish → 6. Update artifacts → 7. Summary

Step 1 — Verify project + locate profile

cd -- '<working_dir>' || { echo "BLOCKED: cannot enter working_dir" >&2; exit 1; }
if [ ! -f power.config.json ] || [ ! -f app.config.js ]; then
  echo "BLOCKED: working_dir is not an initialized app" >&2
  exit 1
fi

Read <working_dir>/power.config.json and require a nonempty environmentId matching the owner's selected environment. Reuse matching complete environment context, or perform this non-persisting lookup from the bound root:

cd -- '<working_dir>' || { echo "BLOCKED: cannot enter working_dir" >&2; exit 1; }
environment_id=$(node -p "require('./power.config.json').environmentId || ''") || {
  echo "BLOCKED: unreadable power.config.json" >&2; exit 1;
}
if [ -z "$environment_id" ] || [ "$environment_id" != "<selected-environment-id>" ]; then
  echo "NEEDS_CONTEXT: selected environment does not match working_dir" >&2
  exit 1
fi
node "${PLUGIN_ROOT}/scripts/resolve-environment.js" "$environment_id" --no-cache --require-tenant

Require the returned environment ID, HTTPS URL, and tenant to match the owner; incomplete or conflicting context returns NEEDS_CONTEXT. Never drop --no-cache --require-tenant, change configuration, or switch auth/environment to recover a failure. Use that exact <envUrl> and <tenantId> for every request.

Profile ID resolution (same priority as /assign-offline-profile Step 1):

SourceUsed when
$ARGUMENTS --profile-id <guid>Explicit selection matching the owner's profile context
<working_dir>/offline-profile.json top-level profileIdDefault for /setup-offline-profile-created projects, matching the owner's profile context
OtherwiseReturn NEEDS_CONTEXT to the owner to select the intended profile; never pick another profile silently

STOP if no profile found: "Run /setup-offline-profile first."

Verify the selected live profile belongs to the selected environment. A conflicting snapshot/argument returns NEEDS_CONTEXT; a 404 or permission failure is BLOCKED, not permission to discover a substitute. Resolve each <itemId> only within this profile and verify its table matches the exact requested logical name.

<working_dir>/power.config.json is intentionally NOT consulted for the profile ID. That file is owned by pa app init. The local profile ID lives in <working_dir>/offline-profile.json only.

Step 2 — Resolve target (what to edit)

Parse $ARGUMENTS:

Flag patternEffect
--rename <new-name>Update profile name
--describe <text>Update profile description
--table <logical-name> --scope <0|1|2>Change one table's recorddistributioncriteria. Combine with --me, --team, --bu to set sub-flags when scope=2.
--table <logical-name> --sync <minutes>Change one table's syncintervalinminutes (range 5–1440)
--table <logical-name> --columns add:col1,col2 remove:col3Add or remove logical names from selectedcolumns. Comma-separated, both add/remove optional.
--table <logical-name> --columns resetReplace selectedcolumns with union of always-include + manifest lookups + screen-grep'd (re-runs the architect's Step 6 union for this table)

If no flags → interactive picker. AskUserQuestion with up-to-4 most likely edits:

  • "Rename profile"
  • "Change a table's scope" → next message asks which table
  • "Change a table's sync frequency" → next message asks which table + value
  • "Edit a table's column list" → next message asks which table + add/remove

Restrict the picker and explicit flags to the owner's exact current scope. --columns reset reads the selected table from <working_dir>/.datamodel-manifest.json or <working_dir>/docs/plan-artifacts/.datamodel-manifest.json and greps only absolute <working_dir>/app/ and <working_dir>/src/ paths. Keep the chosen absolute manifest path for the whole invocation; no other-app fallback. Changing the table, profile, or operation requires returning NEEDS_CONTEXT to the owner, not expanding an existing approval.

Step 3 — Show current vs proposed (single gate)

GET the current item state from Dataverse for any tables being edited:

cd -- '<working_dir>' || { echo "BLOCKED: cannot enter working_dir" >&2; exit 1; }
node "${PLUGIN_ROOT}/scripts/dataverse-request.js" "<envUrl>" GET \
  "mobileofflineprofileitems(<itemId>)?\$select=name,recorddistributioncriteria,recordsownedbyme,recordsownedbymyteam,recordsownedbymybusinessunit,syncintervalinminutes,selectedcolumns" \
  --tenant-id "<tenantId>"

Render a current/proposed diff:

Profile  : <name> (<id>)
Editing  : <table-logical-name> item

                Current     →    Proposed
Scope         : Org+me     →    All records
Sync (min)    : 10         →    30
Columns       : 14         →    16  (add: chnl_notes, chnl_actual_visit_date)

AskUserQuestion: "Apply this change? [Apply / Cancel]"

If Apply → continue. If Cancel → STOP.

Step 4 — PATCH

Build the PATCH body with only the fields that changed:

cd -- '<working_dir>' || { echo "BLOCKED: cannot enter working_dir" >&2; exit 1; }
node "${PLUGIN_ROOT}/scripts/dataverse-request.js" "<envUrl>" PATCH \
  "mobileofflineprofileitems(<itemId>)" --tenant-id "<tenantId>" \
  --body '{
    "recorddistributioncriteria": <new>,
    "recordsownedbyme": <new-bool>,
    "syncintervalinminutes": <new>,
    "selectedcolumns": "{\"Columns\":[...]}"
  }'

For profile-level edits (name / description):

cd -- '<working_dir>' || { echo "BLOCKED: cannot enter working_dir" >&2; exit 1; }
node "${PLUGIN_ROOT}/scripts/dataverse-request.js" "<envUrl>" PATCH \
  "mobileofflineprofiles(<profileId>)" --tenant-id "<tenantId>" \
  --body '{"name": "...", "description": "..."}'

Check returned HTTP statuses as well as process exit. A 403/PrivilegeCheckFailed is BLOCKED; do not change identity or profile to bypass permissions. Preserve every field and table outside the approved delta.

Step 5 — Publish

Use the targeted PublishXml recipe from shared/references/dataverse-offline-api.md §9:

cd -- '<working_dir>' || { echo "BLOCKED: cannot enter working_dir" >&2; exit 1; }
node "${PLUGIN_ROOT}/scripts/dataverse-request.js" "<envUrl>" POST \
  "PublishXml" --tenant-id "<tenantId>" --body '{
    "ParameterXml": "<publish><mobileofflineprofiles><mobileofflineprofile><profileId></mobileofflineprofile></mobileofflineprofiles></publish>"
  }'

Publishes only this profile, not the entire org's customizations. Avoids the 429 rate-limit storms on shared envs that the legacy PublishAllXml triggered.

On 400 / 0x80071141 "circular relationship" — same handling as /setup-offline-profile Step 8 (parse cycle path, prompt user to drop one association, retry). Fallback to PublishAllXml only if PublishXml returns an unexpected error other than the cycle case and the owner/user approves that broader publication; a profile-scoped approval does not cover it.

All linked recovery commands, re-GET verification, and retries must bind the same root and pass the same environment/tenant/profile/table/item context. Never retry on a different profile or bypass a permission failure. Confirm publication before Step 6.

If a linked recipe requires another offline helper, return the proposed change to the owner for approval first, then read the helper at its absolute ${PLUGIN_ROOT}/skills/<helper>/SKILL.md path. Forward:

Arguments: --working-dir '<working_dir>' --profile-id "<profileId>" --table "<table>" <exact-approved-change-flags>
Context: MOBILE_APP_ORCHESTRATING=1, orchestrator, working_dir, phase, approved_scope,
         selected environment ID/URL/tenant, profile ID, table allowlist, item ID,
         supplied answers, --plan-only (if present)

The recipient must enforce the same canonical root contract. Missing context returns NEEDS_CONTEXT; do not fall back to direct invocation or rediscovery.

Step 6 — Update artifacts

Re-read the changed item(s) in the same profile/environment and rewrite the matching entry in <working_dir>/offline-profile.json. When the edit changed a table's columns (--columns add:/remove:/reset), also refresh that entry's schemaColumns to the table's current full column set from the bound absolute manifest. This re-baselines the schema-reconciliation marker for the next ${PLUGIN_ROOT}/scripts/offline-profile-delta.js run; leave schemaColumns untouched for scope/sync/rename-only edits. Any delta check uses --project-root '<working_dir>' in its own guarded call. See offline-profile-reconciliation.md.

Append a one-line entry to <working_dir>/memory-bank.md ## Offline profile block:

edits:
  - { at: 2026-05-19T..., field: 'chnl_storevisit.syncintervalinminutes', from: 5, to: 10 }

Step 7 — Summary

✓ Edited profile.

  Field : chnl_storevisit.syncintervalinminutes
  From  : 5 min
  To    : 10 min
  Published: 2026-05-19T...

offline-profile.json + memory-bank.md updated.

Status code (final line)

  • DONE — change applied and published
  • DONE_WITH_CONCERNS: <list> — change applied but with caveats (publish timeout-then-success, etc.)
  • NEEDS_CONTEXT: <missing> — couldn't resolve target table or profile
  • BLOCKED: <reason> — auth or PATCH failure

What this skill does NOT do

  • Add a NEW table to the profile → use /add-table-to-offline-profile
  • Add a new association (relationship inclusion) → blocked on v0.2 selectedrelationshipsschema work; use maker portal in the meantime
  • Delete the whole profile → manual DELETE /mobileofflineprofiles(<id>) (cascade-deletes items + associations); see dataverse-offline-api.md §11
  • Migrate the profile between environments → use CloneMobileOfflineProfile action (v0.5 work)

microsoft의 다른 스킬

oss-growth
microsoft
OSS 성장 해커 페르소나
agent-framework-azure-ai-py
microsoft
Microsoft Agent Framework Python SDK(agent-framework-azure-ai)를 사용하여 Azure AI Foundry 에이전트를 구축합니다. AzureAIAgentsProvider로 지속적 에이전트를 만들 때, 호스팅 도구(코드 인터프리터, 파일 검색, 웹 검색)를 사용할 때, MCP 서버를 통합할 때, 대화 스레드를 관리할 때, 또는 스트리밍 응답을 구현할 때 사용합니다. 함수 도구, 구조화된 출력, 다중 도구 에이전트를 다룹니다.
development
airunway-aks-setup
microsoft
AKS에서 AI Runway 설정 — 빈 클러스터에서 실행 중인 모델까지. 클러스터 검증, 컨트롤러 설치, GPU 평가, 공급자 설정, 첫 배포를 다룹니다. 시기: "AI Runway 설정", "AKS 클러스터 온보딩", "AI Runway 설치", "airunway 설정", "AKS에 모델 배포", "AKS에서 GPU 추론", "AKS에서 KAITO 설정", "AKS에서 LLM 실행", "AKS에서 vLLM", "AKS에서 모델 서빙 설정", "AI Runway 컨트롤러".
devops
appinsights-instrumentation
microsoft
Azure Application Insights로 웹앱을 계측하기 위한 지침입니다. 원격 분석 패턴, SDK 설정, 구성 참조를 제공합니다. WHEN: 앱 계측 방법, App Insights SDK, 원격 분석 패턴, App Insights란 무엇인가, Application Insights 지침, 계측 예시, APM 모범 사례.
devops
applicationinsights-web-ts
microsoft
브라우저/웹 앱을 Application Insights JavaScript SDK(@microsoft/applicationinsights-web)로 계측합니다. Real User Monitoring(RUM) — 페이지 뷰, 클릭, AJAX/fetch 종속성, 예외, 사용자 지정 이벤트, 백엔드 OpenTelemetry 트레이스와 상관관계가 있는 브라우저 측 GenAI 에이전트 트레이스에 사용합니다. SDK Loader Script 및 npm 설정, 프레임워크 확장(React, React Native, Angular), Click Analytics, 텔레메트리 이니셜라이저, 브라우저에서 생성된 에이전트/도구/모델 스팬에 대한 OTel GenAI 의미론적 규칙을 다룹니다.
devops
azure-ai-anomalydetector-java
microsoft
Azure AI Anomaly Detector SDK for Java로 이상 탐지 애플리케이션을 구축하세요. 단변량/다변량 이상 탐지, 시계열 분석 또는 AI 기반 모니터링을 구현할 때 사용하세요.
development
azure-ai-language-conversations-py
microsoft
azure-ai-language-conversations Python SDK를 사용하여 대화형 언어 이해(CLU)를 구현합니다. ConversationAnalysisClient로 대화 의도와 엔터티를 분석하거나, NLP 기능을 구축하거나, 애플리케이션에 언어 이해를 통합할 때 사용합니다.
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 for Python. ML 작업 영역, 작업, 모델, 데이터 세트, 컴퓨팅 및 파이프라인에 사용합니다. 트리거: "azure-ai-ml", "MLClient", "workspace", "model registry", "training jobs", "datasets".
development