메뉴
Query OWS model data — resolves process prefix, discovers fields, builds and runs TQL. Use when asked about tickets, CMs, INCs, BOTs, or any process data.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
Validate CIT account-upgrade tickets sitting in the AUTIN approval queue. Reports APPROVE / REJECT / NEEDS_REVIEW per ticket based on attachment rules derived from the requester's USER_EXTRA profile, then awaits user confirmation to execute approval.
Trace and debug end-to-end flows in OWS — page → service, service → service, and page → page navigation. TRIGGER when the user asks "why is X failing", "what's wrong with this page/service", "trace the flow of Y", "audit the call chain", or describes broken behavior in an OWS app. Inspects flow steps, RunScript bodies, input/output schemas, log traces, and reference graph; identifies bad code patterns (silent catch, hardcoded IDs, missing validation), schema mismatches (caller passes X but service expects Y), and dead branches; reports findings with concrete file/step/line locations and recommended fixes.
Capture a reusable OWS/ADC platform finding into the knowledge vault. Use when you confirm a non-obvious, reusable fact about the OWS platform (an API path, a TQL quirk, an auth detail, a service contract) that future sessions should know — so it's recorded once and read first via get_help_home / search_help instead of re-derived.
Decision tree for choosing the right OWS MCP tool. TRIGGER whenever the user asks about OWS, GDE, Huawei tickets, or anything that requires calling an `mcp__ows-gde__*` tool — including queries about tickets (CM, INC, BOT, TTS, CTT), pages, services, models, triggers, scripts, processes, modules, projects, logs, or Studio elements. Use BEFORE making the first MCP call to pick the leanest tool variant and avoid over-fetching.
Checklist and patterns for building a new OWS page element. Use when asked to create or design a page.
Checklist and patterns for building a new OWS service element. Use when asked to create or design a service.
| name | ows-query |
| description | Query OWS model data — resolves process prefix, discovers fields, builds and runs TQL. Use when asked about tickets, CMs, INCs, BOTs, or any process data. |
Resolve the process — use resolve_process_by_prefix with the ticket prefix (CM, INC, BOT, TTS, etc.) to get tickets_uri.
Discover fields — if the query needs a field you're unsure about, use get_model_fields(asset_uri=tickets_uri) to find the exact field name. Field names are snake_case and may differ from their display labels (e.g. createtime not createTime).
Build TQL — use this syntax:
select * from "<tickets_uri>" as t where <conditions> order by t.<field> desc limit <N>
as t)t.<fieldname> for ordering and filteringRun the query — use query_model_data(tenant, tql). It validates before executing.
Present results — extract and show only the relevant fields in a table. Don't dump all 200+ fields.
| Field | Meaning |
|---|---|
createtime | When the ticket was created |
lastupdatetime | Last update |
ticketstatus | running / closed / aborted |
task_status | dispatched / accepted / completed |
orderidview | Human-readable ticket ID |
title | Ticket title |
region | Geographic region |
site_id_name | Site ID |
assign_to_fme_full | Assigned engineer name |
fault_type | Wireless / Transmission / Power |
tenant: "prod" for real data unless the user says testbedget_model_fieldsorder by t.createtime desc limit Nt.createtime >= '2026-05-20 00:00:00'