메뉴
Compute month-end variance — driver decomposition (volume/price/mix/one-off heuristic), anomaly flags, and a matplotlib bar chart of variance % by cost centre saved to /tmp for download.
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
| name | variance-analysis |
| description | Compute month-end variance — driver decomposition (volume/price/mix/one-off heuristic), anomaly flags, and a matplotlib bar chart of variance % by cost centre saved to /tmp for download. |
| enabled | true |
Use this skill whenever the controller asks for a variance review, a chart, or any "is this number reasonable?" sanity check. This is the computation surface — it uses code_interpreter against the numbers returned by sap-s4. Do NOT do mental math on financial figures; always send the numbers through this skill.
sap_get_variance_analysis for the requested period.code_interpreter with the analysis script below — DO NOT retype the numbers.investigate, watch, normal)/tmp/variance-<period>.png/tmp/variance-<period>.csv for the close packThe full script is in scripts/variance_analysis.py in this skill's directory. Read it once with the file_read tool, then invoke it via code_interpreter — DO NOT inline the script body in your response.
The script signature:
python /app/use-cases/finance-close/skills/variance-analysis/scripts/variance_analysis.py \
--period 2026-05 \
--rows-json '<the JSON returned by sap_get_variance_analysis>' \
--out-dir /tmp
Outputs to stdout: the per-bucket tables. Outputs to disk: /tmp/variance-2026-05.png and /tmp/variance-2026-05.csv.
The script also classifies each investigate row with one of:
gl_account.type == "OpEx" and prior-year run-rate close, but unit-rate moved (Sentinel-style licence uplift)sap_list_journal_entriesThis is a suggestion to seed owner commentary, not a determination. Always present it as Suggested driver: … and let the human override.
sap_get_variance_analysis inline; don't build the whole chart.code_interpreter directly without this skill.Read the plant-floor Azure IoT surface — device twins (spindles, conveyors, robotic arms, vision systems), time-series telemetry with pre-annotated alarms, downtime events with cause codes and lot ids, and per-line OEE rollups (Availability × Performance × Quality).
Share generated files (PDFs, charts, CSVs) with Frank by writing them to /tmp and referencing the absolute path. The frontend auto-converts /tmp paths into download links.
Build the downloadable incident brief PDF — cover + KPI strip + OEE chart + downtime timeline + root-cause narrative + supplier/material signals + recommended actions + sign-off, per SOP §7.1. Renders via the inline HTML template + Playwright.
Microsoft 365 surface — READ Frank's mailbox / Wes's calendar / OneDrive (shift reports, SOPs) AND a strict H-I-T-L workflow to draft and send an email (e.g. brief Wes on the morning incident).
Compute and plot OEE — a 7-day trend chart with target/threshold lines, an overlay of telemetry alarms and downtime stops, plus a flat CSV export. Saves /tmp/oee-trend-<plant>-line-<N>-<date>.png + .csv.
The Olympus Cleveland-plant Standard Operating Procedure — alarm thresholds, escalation matrix, supplier quality-hold rules, OEE definitions, incident-brief template requirements. Cite the section number when answering.