Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add vigneshbarani24/sap-superpowers --skill cutover-planninggit clone --depth 1 https://github.com/vigneshbarani24/sap-superpowersWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/vigneshbarani24/sap-superpowers/cutover-planning)<a href="https://agentmods.dev/skills/vigneshbarani24/sap-superpowers/cutover-planning"><img src="https://agentmods.dev/badge/skills/vigneshbarani24/sap-superpowers/cutover-planning/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/vigneshbarani24/sap-superpowers/cutover-planning"><img src="https://agentmods.dev/badge/skills/vigneshbarani24/sap-superpowers/cutover-planning.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00068 | $0.03569 |
| Opus 5 | $0.00034 | $0.01784 |
| Sonnet 5 | $0.00014 | $0.00714 |
| Haiku 4.5 | $0.00007 | $0.00357 |
Grade A, and why
cutover-planning scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 191 lines — stays where its author put it; the contents beside it link to each section on GitHub.
SAP Cutover Planning and Execution
This skill enforces a complete, rehearsed, and evidence-based cutover plan — preventing undocumented tasks, missed dependencies, unconfirmed interface partners, and go-live without a tested rollback procedure.
Iron Laws
- NEVER EXECUTE CUTOVER WITHOUT AT LEAST ONE REHEARSAL. A cutover plan untested in a dress rehearsal is a hypothesis, not a plan. Surprises on go-live weekend are not recoverable.
- NEVER SKIP THE GO/NO-GO CHECKPOINT. Go/No-Go is a formal gate with named decision-makers and documented criteria. "We'll see how things look" is not a Go/No-Go process.
- ALWAYS HAVE A DOCUMENTED AND TESTED ROLLBACK PLAN. If rollback has not been rehearsed, it does not exist. The rollback plan must include time limits, decision authority, and step-by-step execution procedures.
- NEVER ASSUME INTERFACE PARTNERS ARE READY WITHOUT WRITTEN CONFIRMATION. Every external system, EDI partner, and third-party interface must provide written confirmation of readiness before cutover begins. Verbal confirmation is not confirmation.
- ALWAYS MAINTAIN A REAL-TIME CUTOVER STATUS BOARD. Every task on the runbook must have an owner, a start time, an end time, and a live status (Not Started / In Progress / Complete / Blocked). No task exists unless it is tracked.
- NEVER EXTEND THE CUTOVER WINDOW WITHOUT INVOKING THE ROLLBACK DECISION PROCESS. If the cutover falls behind the buffer threshold, rollback criteria must be formally evaluated — not silently extended.
- SEQUENCE IS NOT OPTIONAL. Technical tasks, functional tasks, and organizational tasks have hard dependencies. Executing them out of sequence causes data corruption, interface failures, and system inconsistency.
Rationalization Table
| Agent Will Try To... | Why It Seems Reasonable | Why It Fails | Counter |
|---|---|---|---|
| Skip dress rehearsal because "we know what we're doing" | "The team is experienced and the plan is solid" | Rehearsals expose timing errors, missing dependencies, and coordination gaps that experience cannot predict. Dress rehearsals routinely add 20-40% to planned duration. | Iron Law 1: No cutover without rehearsal. If time is short, run an abbreviated rehearsal — but run it. |
| Omit rollback procedures because "we won't need them" | "The go-live will succeed — why plan for failure?" | Industry data: 15-20% of SAP go-lives require partial or full rollback. No rollback plan means rollback is uncontrolled and potentially catastrophic. | Iron Law 3: Rollback plan and time limit are mandatory deliverables. |
| Accept verbal confirmation from interface partners | "They said they'd be ready on the call" | Verbal commitments are forgotten, misunderstood, and legally unenforceable. Integration failures on go-live weekend are the most common single cause of cutover extension. | Iron Law 4: Written confirmation required. Email is acceptable; phone call notes are not. |
| Merge technical and functional cutover tasks into a single undifferentiated list | "It's simpler to track everything together" | Technical and functional tasks have different owners, different dependencies, and different rollback implications. Merging them obscures critical-path analysis. | Checklist Step 2: Tasks must be categorized by type with ownership explicitly assigned. |
| Plan cutover without buffer time | "The sequence is tight but achievable" | Every SAP cutover has unexpected issues. Without buffer, any delay triggers a rollback decision. A 20-30% time buffer is not padding — it is risk management. | Checklist Step 4: Buffer time is mandatory and must be calculated explicitly. |
| Treat go/no-go as a team consensus | "Everyone agrees we should proceed" | Consensus is not accountability. Go/No-Go requires named decision-makers who bear personal accountability for the decision. | Iron Law 2: Named decision-makers documented in advance. |
| Use a spreadsheet cutover plan without a real-time status mechanism | "Everyone can see the Excel" | Static spreadsheets become stale within hours of cutover start. Real-time visibility requires a live tracking tool or dedicated war room status board. | Iron Law 5: Real-time cutover status board is mandatory. |
| Skip DNS and RFC cutover steps because "IT will handle it" | "That's an infrastructure task, not my concern" | RFC connections, DNS changes, and system aliases must be sequenced precisely with functional activation. Missequencing causes connection failures that appear as functional errors. | Checklist Step 5: System landscape changes are explicit cutover tasks with dependencies documented. |
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 191 lines · 68 tokens per session scan A c29717f5e061
cutover-planning is a skill published in the GitHub repository vigneshbarani24/sap-superpowers (9 stars, last pushed 18d ago), licensed MIT. It adds 68 tokens to every session and 3,569 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
sap-expert
Expert in SAP ERP systems, ABAP programming, SAP HANA, S/4HANA, Fiori applications, and SAP integration patterns including OData, RFC, and IDoc. Use when the user mentions ERP, enterprise, business apps, ABAP, HANA, or S/4HANA, or when the task involves SAP Ecosystem, ABAP Development, Integration Technologies, or…
migrate-from-openclaw
Migrate from OpenClaw to NanoClaw v2. Detects an existing OpenClaw installation, extracts identity, channel credentials, scheduled tasks, and other config, then guides interactive migration. Triggers on "migrate from openclaw", "openclaw migration", "import from openclaw".
slack-agent-flow
Let an existing Slack agent create new agents that arrive as their own Slack bots — provisioned app, operator DM, and a shared three-way room, hot-started without a host restart.
add-dial-tool
Give chosen NanoClaw agents a real phone number as a container tool — the dial CLI baked into the agent image plus OneCLI credential injection for api.getdial.ai, scoped per agent, so the agents you pick can send SMS, place AI voice calls, and receive verification codes from inside the sandbox. Independent of the Dial…
add-signal
Add Signal channel integration via signal-cli device-link. Native adapter — no Chat SDK bridge.
add-deltachat
Add DeltaChat channel integration via @deltachat/stdio-rpc-server. Native adapter — no Chat SDK bridge. Email-based messaging with end-to-end encryption.