Daily Checks Audit Skill

This skill enables senior consultants to validate SAP daily monitoring reports performed by offshore teams without logging into the SAP systems.

Overview

The skill analyzes Excel-based daily monitoring reports and generates an audit report that identifies:

• •❌ Critical Issues: Negative (N) responses without justification
• •⚠️ Warnings: Threshold violations, missing explanations for anomalies
• •✅ Passed Checks: Validations that meet all criteria

Supported Customers

Prerequisites

• •Python 3.8+ installed
• •openpyxl package: pip install openpyxl

Usage

Step 1: Place the Excel File

Place the daily monitoring Excel file in the project directory:

c:\GenAI\daily_checks_audit\

Step 2: Run the Audit

cd c:\GenAI\daily_checks_audit
python .agent/skills/daily-checks-audit/scripts/audit_checks.py <excel_filename>

Examples:

# TBS customer
python .agent/skills/daily-checks-audit/scripts/audit_checks.py TBS_DAILY_MONITORING_20_JAN_2026.xlsx

# SONOCO customer (EVIOSYS files)
python .agent/skills/daily-checks-audit/scripts/audit_checks.py CENTIQ_SAP_EVIOSYS_DAILY_MONITORING_January_2026_01_20.xlsx

Step 3: Review the Report

The script generates a markdown report: audit_report_<timestamp>.md

The report includes:

• •Executive Summary: Total issues (critical/warnings)
• •Check Metadata: Date, time, performer for each system
• •Per-System Findings: Detailed issues by system/sheet
• •Recommendations: Action items for critical issues

Validation Rules

Critical Issues (Require Immediate Action)

Warnings (Review Required)

Customer-Specific Thresholds

Each customer has different "normal" ranges based on historical data:

Adding a New Customer

1. •

   Create config file: Copy an existing config and modify:

   code

   configs/<CUSTOMER>_config.json
   

2. •

   Update detection: Edit scripts/audit_checks.py, add to detect_customer():

   python

   for prefix in ['TBS', 'BSW', 'COREX', 'SONOCO', 'NEWCUSTOMER']:
   

3. •

   Analyze historical data: Run analysis on 5-10 files to determine appropriate thresholds

Output Example

[AUDIT] Auditing: TBS_DAILY_MONITORING_20_JAN_2026.xlsx
[CONFIG] Loaded TBS customer configuration
[INFO] Found 8 system sheets

[RESULTS] Audit Results:
   [!] Critical: 4
   [?] Warnings: 1
   Total Issues: 5

[OK] Report saved to: audit_report_20260120_113157.md

Troubleshooting

Report Interpretation Guide

Critical Issues

• •Must be addressed immediately
• •Typically indicates checks were not performed properly
• •Example: "Row 77: Negative (N) response without justification"

Warnings

• •Review required but may be acceptable
• •Often indicates values outside normal range
• •Example: "Old locks present (16) without explanation"

AI Agent Usage

This skill is designed to be invoked by AI agents (e.g., Gemini, Claude). When a user mentions daily checks or monitoring audits, the agent should use this skill.

Trigger Phrases

The agent should activate this skill when the user says things like:

• •"Audit the daily checks for [filename]"
• •"Validate the daily monitoring report"
• •"Check if the offshore team did the daily checks correctly"
• •"Review [customer] daily monitoring file"
• •"@[filename.xlsx] - audit this daily check file"

Agent Workflow

1. •Identify the Excel file from user's message or @mention
2. •Run the audit script:
   powershell

   cd c:\GenAI\daily_checks_audit
   python .agent/skills/daily-checks-audit/scripts/audit_checks.py <filename>
   

3. •Read the generated report (audit_report_*.md)
4. •Summarize findings for the user

Expected Agent Response Format

After running the audit, the agent should provide:

**Audit Summary for [Customer] - [Date]**

📊 **Results**: X critical issues, Y warnings

**Critical Issues Requiring Action:**
1. [System] Row [N]: [Issue description]
2. ...

**Warnings to Review:**
1. [System] Row [N]: [Issue description]

**Recommendation**: [Follow-up action if needed]

Example Agent Interaction

User: "Audit TBS_DAILY_MONITORING_20_JAN_2026.xlsx"

Agent Actions:

1. •Runs: python .agent/skills/daily-checks-audit/scripts/audit_checks.py TBS_DAILY_MONITORING_20_JAN_2026.xlsx
2. •Reads generated report
3. •Responds:

Audit Summary for TBS - January 20, 2026

📊 Results: 4 critical issues, 1 warning

Critical Issues:

1. •CRP Row 77: N response without justification (Response times check)
2. •BIP Row 75: N response without justification
3. •BOP Row 7: N response without justification
4. •JIP Row 8: N response without justification (system OK check)

Warnings:

1. •ERP Row 66: 16 old locks without explanation

Recommendation: Follow up with Preeti Zirwal and Shivam Srivastava regarding the missing justifications.

Handling Multiple Files

If the user provides multiple files or asks for batch processing:

1. •Run the audit script on each file sequentially
2. •Summarize findings across all files
3. •Highlight any patterns or recurring issues

Error Handling

File Structure

.agent/skills/daily-checks-audit/
├── SKILL.md                    # This instruction file
├── scripts/
│   └── audit_checks.py         # Main audit script
└── configs/
    ├── BSW_config.json         # BSW thresholds
    ├── TBS_config.json         # TBS thresholds
    ├── COREX_config.json       # COREX thresholds
    └── SONOCO_config.json      # SONOCO thresholds