All skills
secondsky avatar

/sap-rpt1

@620a19a
by Eddiesecondsky/sap-skills456 stars
120

SAP-RPT-1-OSS local tabular prediction workflows for FI/CO prototype datasets. Use when preparing SAP finance CSV exports for classification or regression experiments with source-verified setup, leakage checks, and governance review.

Use this Skill: https://skilld.dev/gh/secondsky/sap-skills/sap-rpt1

This session only. Nothing lands on disk.

referencesfico-use-cases.md

≈1.6k tokens on demand. Your agent reads this file only when SKILL.md points to it.

FI/CO Use Cases for SAP-RPT-1-OSS

Sources

Derived from public SAP-RPT-1-OSS source/model review and general SAP FI/CO table knowledge. Validate all table and field choices against the target SAP release and customer configuration before using real data.

Scope

This reference provides FI/CO tabular prediction patterns for prototype and research workflows. It is not an extractor, production scoring service, hosted API client, or compliance approval workflow.

Use-Case Matrix

Area Use case Typical target Common S/4HANA starting points ECC fallback starting points Leakage warning
FI-AR Payment default / late payment paid_late, days_late, default_flag ACDOCA, customer master, clearing/payment extracts BSID, BSAD, BKPF, BSEG, KNA1 Do not use fields updated after due date or clearing.
FI-AP Payment block / payment timing blocked_at_payrun, late_payment ACDOCA, vendor master, payment proposal extracts BSIK, BSAK, BKPF, BSEG, LFA1 Do not use later payment-run decisions as features.
FI-AP Cash discount leakage discount_lost, discount_amount_lost ACDOCA, payment terms, baseline/due-date fields BSIK, BSAK, BKPF, BSEG, LFB1 Do not use final clearing/payment date if predicting before payment.
FI-GL Journal anomaly manual_review_flag, reversal_flag, outlier_flag ACDOCA, journal header metadata BKPF, BSEG, SKA1, SKB1 Do not train on later audit decisions unless time-split.
CO Cost center overrun overrun_flag, overrun_amount Universal journal actuals plus plan/budget extracts COEP, COSS, COSP, CSKS, CSKT Do not use actuals posted after the forecast date.
CO Internal order overrun overrun_flag, remaining_budget_risk Universal journal actuals plus order/budget extracts COEP, AUFK, BPGE, BPJA Do not use final settlement outcomes as pre-period features.
CO/PS Project budget overrun overrun_flag, estimate_at_completion_error Universal journal actuals plus project/budget extracts PRPS, PROJ, COEP, BPGE, BPJA Do not use later milestone or settlement status.
CO-PA Profitability / margin anomaly margin_anomaly_flag, low_margin_flag Margin/profitability extracts plus ACDOCA CE1*, CE2*, CE3*, CE4* by operating concern Do not use final period adjustments for earlier prediction points.
Treasury/FI Cash forecast risk cash_shortfall_flag, forecast_error AR/AP open items, bank/cash extracts AR/AP open item tables plus bank statement extracts Do not use actual cash movements after forecast date.
FI-AR Dispute risk dispute_created, dispute_escalated AR open items plus dispute extracts AR open items plus FSCM/dispute extracts if available Do not use dispute resolution data as pre-dispute features.
FI-AR/Credit Credit risk credit_block, default_flag, risk_class_change AR history plus credit exposure extracts AR open/cleared items plus credit master/exposure extracts Do not use future block decisions or collection outcomes.

Detailed Recipe: FI-AR Payment Default

Prediction Point

Use invoice state as of a defined date before or on due date. State whether the score is created at posting, before due date, after due date, or before collections review.

Candidate Target

  • paid_late
  • days_late
  • default_flag

Candidate Features

  • Company code.
  • Customer account group or risk segment.
  • Payment terms.
  • Document type.
  • Baseline date bucket.
  • Due date bucket.
  • Invoice amount bucket.
  • Currency-normalized amount.
  • Historical customer delay rate calculated only from prior invoices.
  • Prior dispute count calculated only before the as-of date.
  • Prior dunning count calculated only before the as-of date.

Leakage-Prone Fields

  • Clearing date after due date.
  • Clearing document.
  • Final payment status.
  • Dunning level assigned after the prediction point.
  • Dispute resolution status.
  • Collection outcome.
  • Write-off indicator created after the prediction point.

Minimum Validation

Use a time-based split. Compare metrics by company code and customer segment. Review false positives with finance owners before treating any score as actionable.

Detailed Recipe: FI-AP Cash Discount Leakage

Prediction Point

Use invoice state at posting time, approval time, or before payment proposal. Do not mix prediction points in the same dataset unless each row has an explicit as-of timestamp.

Candidate Target

  • discount_lost
  • discount_amount_lost

Candidate Features

  • Vendor account group or risk segment.
  • Payment terms.
  • Baseline date bucket.
  • Discount window days.
  • Invoice approval age.
  • Amount bucket.
  • Company code.
  • Purchasing group if available before the prediction point.
  • Historical vendor processing delay calculated only from prior invoices.

Leakage-Prone Fields

  • Actual clearing date if prediction is before payment.
  • Payment run result.
  • Final cash discount taken or lost amount.
  • Later payment block changes.
  • Reversal or correction postings after prediction point.

Minimum Validation

Use historical periods for training and later periods for validation. Review results by company code, vendor group, and payment terms.

Detailed Recipe: FI-GL Journal Anomaly

Prediction Point

Use journal state at posting time or before period close review. Record whether the target is rule-generated, auditor-labeled, or generated from later reversals.

Candidate Target

  • manual_review_flag
  • reversal_flag
  • outlier_flag

Candidate Features

  • Company code.
  • Ledger.
  • Fiscal period.
  • Document type.
  • Posting key pattern.
  • Account group.
  • Amount bucket.
  • Manual vs interface origin if available at posting time.
  • Posting time and posting day pattern.
  • Sanitized text/category features only when approved.

Leakage-Prone Fields

  • Later reversal document.
  • Audit decision.
  • Manual investigation notes.
  • Period-close adjustment flags created after posting.
  • Workflow approval outcome if not known at prediction point.

Minimum Validation

Use time-based validation and review by document type, account group, posting origin, and company code. Treat scores as review prioritization only.

Source: SKILL.md on GitHub

1 warning1mo3 checks · Risk SAFE
  • Gen Agent Trust Hub1mo

    This skill provides a secure workflow for analyzing SAP Finance (FI/CO) data using local Python scripts and models from a well-known vendor. It incorporates data privacy checks, governance guidelines, and relies on reputable sources for its functionality.

  • Socket1mo

    No alerts

  • Snyk1mo

    Risk: MEDIUM · 1 issue

Signed by skilld at 620a19a. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 2 weeks ago.

Activeupdated 2 months ago
What it can do
Reads files Runs commands
All 2 allowed tools
ReadBash
Other metadata
metadata
{
  "maintainer": "Eduard Jiglau",
  "maintainer_email": "hello@sap-ai-skills.com",
  "website": "https://sap-ai-skills.com",
  "version": "2.4.1",
  "model_source": "SAP/sap-rpt-1-oss",
  "python_version": "3.11",
  "last_verified": "2026-06-18",
  "requires_huggingface_login": true,
  "verification_scope": "public source/model/product-page review only"
}

README badge

README badge for secondsky/sap-skills/sap-rpt1