SAP FICO Expert — OpenClaw Skill
| Property | Value |
by chanfouricc · published 2026-03-22
$ claw add gh:chanfouricc/chanfouricc-sap-fico-consultant# SAP FICO Expert — OpenClaw Skill
📋 Overview
| Property | Value |
|----------|-------|
| **Name** | `sap-fico-expert` |
| **Version** | 1.0.0 |
| **Platform** | OpenClaw (Telegram / Social platforms) |
| **Target LLM** | DeepSeek Chat (explanations) / DeepSeek Coder (ABAP) |
| **Language** | French technical SAP terminology |
| **Author** | @chanfouricc |
🎯 Objective
Transform any OpenClaw bot into a **Senior SAP Finance & Controlling consultant** capable of answering configuration, troubleshooting, cross-module integration, and S/4HANA migration questions with production-grade accuracy.
📁 Skill Structure
sap-fico-skill/
├── SKILL.md # This documentation
├── skill.json # OpenClaw skill configuration
├── system_prompt.md # Full system prompt (to inject)
├── examples.json # Calibrated few-shot Q&A examples
└── reference/
├── tcodes_index.md # T-code index by domain
├── tables_index.md # Critical SAP tables index
└── error_codes.md # Common FI/CO error messages🚀 Installation on OpenClaw
1. Copy the skill to your VPS
scp -r sap-fico-skill/ user@vps:/opt/openclaw/skills/2. Register the skill in OpenClaw
Add to your OpenClaw configuration (`config.json` or equivalent):
{
"skills": {
"sap-fico-expert": {
"enabled": true,
"trigger": "auto",
"keywords": ["SAP", "FICO", "FI-GL", "FI-AP", "FI-AR", "FI-AA", "CO-CCA", "CO-PA", "CO-PC", "ACDOCA", "S/4HANA", "T-code", "BKPF", "BSEG", "OB52", "FB01", "KS01"],
"system_prompt_path": "skills/sap-fico-expert/system_prompt.md",
"examples_path": "skills/sap-fico-expert/examples.json",
"model_override": "deepseek-chat",
"parameters": {
"max_tokens": 600,
"temperature": 0.25,
"presence_penalty": 0.1
}
}
}
}3. Trigger activation
The skill activates automatically when a message contains an SAP keyword. Alternatively, users can force activation with:
/skill sap-fico-expert🔧 Recommended Settings
| Parameter | Value | Rationale |
|-----------|-------|-----------|
| `temperature` | 0.25 | Deterministic — no creativity on SAP config |
| `max_tokens` | 600 | Increased from initial spec (450 too short for complex questions) |
| `presence_penalty` | 0.1 | Slight lexical diversity without drift |
| `model` | `deepseek-chat` | Default; switches to `deepseek-coder` when ABAP detected |
> ⚠️ **Note**: The original 450 token limit was raised to 600. Responses involving cross-module integrations or S/4HANA migrations need more space to remain actionable.
📊 Coverage
Tier 1 — Core expertise (immediate answers)
Tier 2 — Advanced expertise (detailed answers)
Tier 3 — Integration expertise (contextual answers)
🧪 Testing
Quick validation questions
| # | Test question | Verify |
|---|--------------|--------|
| 1 | "How to configure document splitting in S/4HANA?" | T-code + tables + steps |
| 2 | "Error F5 025 when posting" | Diagnosis + solution + config |
| 3 | "Difference between assessment and distribution in CO?" | Explanation + T-codes + use cases |
| 4 | "How does GR/IR clearing work?" | Process + FI-MM integration + T-codes |
| 5 | "Migrate FI-AA to New Asset Accounting" | Steps + tables + S/4HANA specifics |
Quality criteria
📝 Changelog
| Version | Date | Changes |
|---------|------|---------|
| 1.0.0 | 2026-02-08 | Initial release — full FI/CO coverage |
More tools from the same signal band
Order food/drinks (点餐) on an Android device paired as an OpenClaw node. Uses in-app menu and cart; add goods, view cart, submit order (demo, no real payment).
Sign plugins, rotate agent credentials without losing identity, and publicly attest to plugin behavior with verifiable claims and authenticated transfers.
The philosophical layer for AI agents. Maps behavior to Spinoza's 48 affects, calculates persistence scores, and generates geometric self-reports. Give your...