SAP Superpowers
SAP Superpowers provides enforced workflows and specialized agents for SAP consulting, enhancing efficiency and accuracy.
SAP Superpowers
The SAP consultant's AI operating system — for Claude Code
What superpowers did for software engineering, SAP Superpowers does for SAP consulting. Not just reference docs — enforced workflows with hard gates, checklists, and decision trees that guide you through complex SAP processes step by step.
Why SAP Superpowers?
Other SAP AI tools give you knowledge. We give you process + agents.
- 10 process skills with hard gates — you literally cannot skip steps
- 6 specialized agents that get dispatched automatically to do focused work
- 16 reference skills covering the full SAP lifecycle (dev + functional + PM + OCM)
- 4 slash commands for instant access to common workflows
- Zero telemetry. Pure Markdown. No API keys. No dependencies.
Quick Start
# 1. Install Claude Code (if you haven't)
npm install -g @anthropic-ai/claude-code
# 2. Inside Claude Code, install SAP Superpowers
/plugin marketplace add vigneshbarani24/sap-superpowers
# 3. Start using it — skills activate automatically
# Or use slash commands:
/sap-debug # Systematic SAP troubleshooting
/sap-estimate # Structured effort estimation
/sap-kickoff # SAP Activate project setup
/sap-review # Clean core code review
No config. No API keys. No telemetry. Just SAP expertise.
What's Inside — 26 Skills
Process Skills — Enforced Workflows (10)
These are rigid, superpowers-style skills with hard gates, numbered checklists, and decision trees. You cannot skip steps.
| Skill | Slash Command | What It Enforces |
|---|---|---|
sap-troubleshooting | /sap-debug | 7-step diagnosis: symptom → layer classification → evidence → SAP Notes → root cause → fix → verify |
sap-estimation | /sap-estimate | Forced decomposition with complexity multipliers, always a range (optimistic/realistic/pessimistic) |
sap-go-live-readiness | — | 10 hard gates with evidence requirements — cannot skip Gate 5 without passing Gate 4 |
sap-code-review | /sap-review | ABAP Cloud tier classification → clean core rules → anti-patterns → severity report |
sap-project-kickoff | /sap-kickoff | SAP Activate Phase 0-1: scope → stakeholders → RACI → governance → risks → timeline |
sap-brainstorming | — | SAP standard first → config possible? → released API? → modification (last resort) |
sap-testing-strategy | — | Testing pyramid: ABAP Unit → integration → UAT → regression, with defect management |
sap-change-management | — | Stakeholder analysis → impact assessment → comms plan → training → adoption tracking |
sap-value-advisory | — | Current state → value drivers → TCO → ROI/NPV/IRR → sensitivity → executive presentation |
using-sap-superpowers | — | Meta-skill: routes to correct skills, auto-activates on session start |
Reference Skills — Domain Knowledge (16)
Deep SAP domain knowledge with transaction codes, SAP Notes, SAP Help Portal links, and version-specific guidance.
Technical
| Skill | Coverage | Key Content |
|---|---|---|
abap-cloud | Clean core, RAP, CDS, released APIs | Tier model, API verification methods, naming conventions |
system-administration | Basis, transports, monitoring | 24+ transaction codes, STMS, SM21, SM50, upgrade planning |
btp-development | BTP architecture, CAP, Fiori Elements | Cloud Foundry vs Kyma, SAP Build, deployment patterns |
sap-integration-suite | CPI, API Management, Cloud Connector | 10 adapter types, iFlow patterns, error handling |
sap-security-grc | Authorization, SoD, GRC, firefighter | Role design patterns, 7 common SoD conflicts, audit trail |
sap-joule | SAP's AI copilot | Capabilities, architecture, ROI concerns, Joule vs external AI |
Functional Modules
| Skill | Module | Key Content |
|---|---|---|
financial-accounting | FI | G/L, AP/AR, asset accounting, closing, universal journal (S/4HANA) |
controlling | CO | Cost centers, internal orders, CO-PA, margin analysis (S/4HANA) |
materials-management | MM | Procurement, inventory, MRP, invoice verification, movement types |
sales-distribution | SD | Order-to-cash, pricing condition technique, ATP, output management |
production-planning | PP | BOM, routing, MRP, production orders, capacity planning |
plant-maintenance | PM | Equipment, work orders, preventive maintenance, notifications |
successfactors | HCM/SF | Employee Central, ECP, S/4HANA integration, module sequencing |
Strategic
| Skill | Coverage | Key Content |
|---|---|---|
s4hana-migration | Migration approaches | Brownfield/greenfield/bluefield decision tree, SUM, custom code remediation |
sap-activate-methodology | Project methodology | 6 phases, deliverables, quality gates, fit-to-standard workshops |
rise-licensing | Commercial models | RISE vs GROW decision matrix, HUoM, FUE, SAPS, BTP credits |
See It In Action
/sap-debug — Never Guess at Fixes Again
You: "I'm getting a MESSAGE_TYPE_X dump in VA01"
Claude: Using sap-troubleshooting skill...
Step 1: Capturing symptom
→ Dump type: MESSAGE_TYPE_X, Transaction: VA01
→ What does ST22 show for exception class and program?
You: "CX_SY_ZERODIVIDE in ZSD_PRICING_EXIT"
Claude: Step 2: Classified as ABAP Layer (short dump)
Step 3: Gathering evidence
→ Custom pricing exit dividing by quantity, but quantity is 0
Step 4: SAP Notes search → Note 2345678 matches
Step 5: Root cause → Missing null check
Step 6: Proposed fix with rollback plan...
What makes it different: 7 mandatory steps. Evidence before fix. No jumping to conclusions.
/sap-estimate — Stop Giving Single-Number Estimates
Claude: Using sap-estimation skill...
| Task | Base | Complexity | Adjusted |
|-------------------|--------|-----------|----------|
| CDS Views | 3 days | 1.5x | 4.5 days |
| RAP Service | 4 days | 1.5x | 6 days |
| Fiori UI | 8 days | 1.5x | 12 days |
Range: 30 days (optimistic) → 38 days (realistic) → 57 days (pessimistic)
What makes it different: Forced decomposition. SAP-specific multipliers. Always a range.
sap-go-live-readiness — 10 Gates, Zero Shortcuts
Claude: HARD GATE: All 10 gates must PASS before declaring readiness.
Gate 1: Functional Testing → PASS ✓
Gate 2: Integration Testing → PASS ✓
Gate 3: UAT Sign-off → FAIL ✗ (2 critical defects open)
BLOCKED: Cannot proceed to Gate 4 until Gate 3 passes.
What makes it different: Hard gates. Can't skip steps. Every gate needs evidence.
Architecture
┌─ Claude Code ──────────────────────────────────────┐
│ │
│ ┌─ SAP Superpowers Plugin ─────────────────────┐ │
│ │ │ │
│ │ hooks/session-start │ │
│ │ → Injects meta-skill on every session │ │
│ │ → Routes to correct skill based on context │ │
│ │ │ │
│ │ skills/ (26 SKILL.md files) │ │
│ │ ├── 10 process skills (hard gates) │ │
│ │ └── 16 reference skills (domain knowledge) │ │
│ │ │ │
│ │ commands/ (4 slash commands) │ │
│ │ agents/ (6 specialized subagents) │ │
│ │ │ │
│ └───────────────────────────────────────────────┘ │
│ │
│ Pure Markdown + shell scripts │
│ No npm install. No API keys. No telemetry. │
│ No dependencies. MIT license. │
└─────────────────────────────────────────────────────┘
Skills auto-activate based on keywords:
- "error", "dump", "not working" →
sap-troubleshooting - "estimate", "effort" →
sap-estimation - "go-live", "cutover" →
sap-go-live-readiness - "business case", "ROI" →
sap-value-advisory - Module-specific topics → corresponding reference skill
Agents — Specialists That Do the Work
Skills guide the process. Agents do focused work. When you trigger a skill, it can automatically dispatch a specialized subagent — a separate AI instance that handles a specific task and reports back.
| Agent | Dispatched By | What It Does |
|---|---|---|
sap-reviewer | sap-code-review | Analyzes ABAP/CDS code against clean core rules, classifies ABAP Cloud tier, flags deprecated APIs |
sap-estimator | sap-estimation | Decomposes work items, applies SAP complexity multipliers, produces range estimates |
sap-migration-analyzer | s4hana-migration | Scans custom code for S/4HANA compatibility, classifies as must-fix/should-fix/nice-to-fix |
sap-test-designer | sap-testing-strategy | Generates test cases from process descriptions — positive, negative, boundary, integration |
sap-value-calculator | sap-value-advisory | Builds ROI/TCO models, calculates NPV/IRR, runs sensitivity analysis across scenarios |
sap-security-auditor | sap-security-grc | Reviews authorization for SoD conflicts, flags over-privileged access (SAP_ALL, broad S_TCODE) |
Who This Is For
| Role | Top Skills |
|---|---|
| Functional Consultants (FI/CO, MM/SD, PP/PM) | Module reference skills, estimation, go-live readiness, value advisory |
| ABAP/BTP Developers | Code review, troubleshooting, ABAP Cloud, BTP development |
| Basis Administrators | System administration, troubleshooting, go-live readiness |
| Solution Architects | Brainstorming, S/4HANA migration, integration suite, Activate methodology |
| Project Managers | Project kickoff, estimation, go-live readiness, Activate methodology |
| Change Management Leads | Change management, testing strategy, value advisory |
What Makes Us Different
Most SAP AI tools give you knowledge — here's how a CDS view works, here's the tcode. We give you process — enforced step-by-step workflows where you can't skip steps, with specialized agents doing focused work.
| Capability | SAP Superpowers |
|---|---|
| Process workflows with hard gates | 10 — no other SAP AI tool has any |
| Specialized subagents | 6 — dispatched automatically during workflows |
| Functional module coverage | FI, CO, MM, SD, PP, PM, SF — not just dev tools |
| Project management | Kickoff, estimation, go-live readiness |
| Change management | Full OCM workflow with adoption tracking |
| Value advisory | TCO, NPV, IRR modeling with sensitivity analysis |
| Migration guidance | Decision trees for brownfield/greenfield/bluefield |
| Clean core code review | ABAP Cloud tier classification + anti-pattern detection |
| Telemetry | None. Zero. No phone home. |
| Cost | Free, MIT licensed |
Contributing
We need SAP practitioners. See CONTRIBUTING.md.
Most wanted:
- Reference skills for modules you know deeply
- Process skills for workflows you follow daily
- SAP Notes references and real-world patterns
- Bug reports and feedback from actual SAP projects
FAQ
Q: Do I need an SAP system? A: No. Skills guide your process and provide knowledge. You use them alongside your SAP system.
Q: Does this replace SAP Joule? A: They're complementary. Joule is great for in-system queries. We're great for cross-cutting consulting methodology. Use both.
Q: Is my data sent anywhere? A: No. Pure Markdown files. No telemetry, no API calls, no data collection. Everything runs locally.
Q: Can I use this with Cursor? A: The hook system supports Cursor. Other platforms may work with manual skill invocation.
Q: How is this different from other SAP AI plugins? A: Others provide knowledge-only plugins for developers. We provide process-driven workflows with hard gates you can't skip, specialized agents, and coverage across the entire SAP lifecycle — including functional, PM, OCM, and value advisory.
Roadmap
- v1.0 — 3 process skills, plugin scaffold, hooks, CI
- v2.0 — 26 skills (10 process + 16 reference), 6 agents, 4 commands (you are here)
- v2.1 — MCP servers for SAP ecosystem integration
- v2.2 — Additional agents and community-requested skills
- v3.0 — Pro tier with advanced deliverable generation
License
MIT — use it, fork it, improve it, share it.
SAP, S/4HANA, ABAP, Fiori, RISE with SAP, SAP Activate, SuccessFactors, and related terms are trademarks of SAP SE. This is an independent community project not affiliated with SAP SE.