Install any skill in seconds. Free to start, no credit card required.
Get Started Free →SAP Datasphere development skill with 3 specialized agents, 5 slash commands, and validation hooks. Use when building data warehouses on SAP BTP, creating analytic models, configuring data flows and replication flows, setting up connections, managing spaces and users, implementing data access controls, using the datasphere CLI, or inspecting authenticated Datasphere browser UI state with Microsoft Edge CDP. Covers Data Builder, Business Builder, analytic models, 40+ connection types, real-time r
.claude/skills/secondsky-sap-datasphere/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 535% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 177% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 242% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 325% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 164% | 0% |
SAP Datasphere is SAP's cloud-native data warehouse solution on SAP Business Technology Platform (BTP). It serves as the data foundation within SAP Business Data Cloud (BDC), SAP's unified data and analytics platform that also includes SAP Analytics Cloud, SAP HANA Cloud, SAP Databricks, and curated data products. See references/business-data-cloud.md for the BDC architecture and how Datasphere fits within it.
This skill provides comprehensive guidance for data acquisition, preparation, modeling, administration, and integration.
Use this skill when:
For Datasphere browser triage, use the shared sap-browser-automation skill for manual in-app authentication, consent-gated authenticated Edge profile copying, fresh loopback CDP startup, target selection, and recovery. The local references/edge-cdp-control.md add-on retains Datasphere-specific boundaries. Use CDP for local read-only inspection of Data Builder, Business Builder, monitoring, connection, catalog, and admin UI state; ask before saving, deploying, deleting, importing, exporting, running task chains, or changing users/roles.
| Component | Purpose | Key Objects | |-----------|---------|-------------| | Data Builder | Data acquisition & preparation | Views, Tables, Flows, Task Chains | | Business Builder | Semantic layer modeling | Business Entities, Fact Models, Consumption Models | | Analytic Model | Analytics-ready structures | Dimensions, Facts, Measures, Hierarchies | | Connections | External data sources | 40+ connection types | | Spaces | Logical data containers | Storage, Users, Objects |
Views:
Tables:
Flows:
Create views visually by dragging sources and adding transformations.
Supported Operations:
Best Practices:
For detailed graphical view operations, see references/graphical-sql-views.md.
Create views using SQL or SQLScript.
sql-- Basic SQL View SELECT customer_id, customer_name, SUM(order_amount) AS total_orders FROM orders GROUP BY customer_id, customer_name
SQLScript Support:
For SQL/SQLScript reference, see references/graphical-sql-views.md.
ETL pipelines for data transformation and loading.
Operators:
Execution:
For data flow details, see references/data-acquisition-preparation.md.
Replicate data from source systems to Datasphere or external targets.
Supported Sources:
Supported Targets:
Load Types:
For replication flow configuration, see references/data-acquisition-preparation.md.
Delta-aware transformations with automatic change propagation.
Key Features:
For transformation flow details, see references/data-acquisition-preparation.md.
Orchestrate multiple tasks in sequence or parallel.
Supported Tasks:
Features:
Create analytics-ready semantic models for SAP Analytics Cloud.
Components:
Creating an Analytic Model:
For detailed modeling guidance, see references/data-modeling.md.
Categorize and filter analytical data.
Types:
Features:
Quantifiable values for analysis.
Types:
Aggregation Types:
For measure configuration, see references/data-modeling.md.
Create business-oriented semantic models.
Objects:
For Business Builder details, see references/data-modeling.md.
SAP Datasphere supports 40+ connection types.
SAP Systems:
Cloud Platforms:
Databases:
Streaming:
Other:
For connection configuration, see references/connectivity.md.
| Feature | Description | |---------|-------------| | Remote Tables | Virtual data access | | Data Flows | ETL transformation | | Replication Flows | Data replication | | Model Import | BW/4HANA model transfer |
Logical containers for data and objects.
Configuration:
Operations:
For space management, see references/administration.md.
Standard Roles:
Scoped Roles:
Authentication:
For user management, see references/administration.md.
Capabilities:
Database Analysis:
For monitoring details, see references/administration.md.
Operations:
Features:
Options:
For monitoring details, see references/data-integration-monitor.md.
The datasphere CLI enables command-line administration and automation.
Installation:
bashnpm install -g @sap/datasphere-cli
Authentication:
bash# Interactive login datasphere config auth login # Service key (CI/CD) datasphere config auth login --service-key-path ./key.json
Core Commands:
| Command | Purpose | |---------|---------| | datasphere spaces list | List all spaces | | datasphere spaces create | Create a space | | datasphere objects export | Export objects | | datasphere objects import | Import objects | | datasphere tasks run | Execute task chains | | datasphere marketplace list | List marketplace products |
CI/CD Integration:
bash# Export and import workflow datasphere objects export --space DEV --output-file package.zip datasphere objects import --space PROD --input-file package.zip --overwrite
For complete CLI reference, see references/cli-commands.md.
Package curated data for internal or external consumption:
Product Components:
Discover and consume published data products:
For complete marketplace guidance, see references/data-products-marketplace.md.
Centralized discovery and governance:
Create Object → Add Metadata → Link Terms → Quality Check → Approve → PublishRoles:
For detailed governance guidance, see references/catalog-governance.md.
Implement row-level security.
Types:
Application:
For security configuration, see references/data-access-security.md.
Move content between tenants.
Methods:
Package Contents:
For transport procedures, see references/content-transport.md.
| Error | Cause | Solution | |-------|-------|----------| | Deployment failed | Circular dependency | Check object dependencies | | Connection timeout | Network/firewall | Verify Cloud Connector/IP allowlist | | Replication stuck | Source lock | Check source system status | | Out of memory | Large view | Enable persistence or partitioning | | Permission denied | Missing role | Verify space membership and privileges |
Core Data Builder:
references/data-acquisition-preparation.md - Data flows, replication flows, transformation flows, and table managementreferences/graphical-sql-views.md - Graphical views, SQL views, E-R models, and intelligent lookupsreferences/data-modeling.md - Business Builder entities, analytic models, dimensions, measures, and hierarchiesConnectivity & Integration:
references/connectivity.md - All 40+ connection types including SAP systems, cloud providers, and streaming platformsreferences/data-integration-monitor.md - Task scheduling, monitoring, real-time replication, and delta mechanismsAdministration & Security:
references/administration.md - Tenant management, space configuration, user roles, and elastic compute nodesreferences/data-access-security.md - Row-level security, DAC configurations, and authorization scenariosreferences/content-transport.md - Package export/import, transport management, and tenant migrationCLI & Automation:
references/cli-commands.md - Complete CLI reference, authentication, CI/CD integration patternsMarketplace & Governance:
references/data-products-marketplace.md - Creating and consuming data products, provider workflows, pricingreferences/catalog-governance.md - Data catalog, glossary, quality rules, lineage, classificationBest Practices & Updates:
references/best-practices-patterns.md - Architecture patterns, naming conventions, performance optimization, checklistsreferences/whats-new-2025.md - 2025 archive: Q1-Q4 2025 features, Generic HTTP, REST API tasks, deprecationsreferences/whats-new-2026.md - 2026 Datasphere/BDC features, including Q1-Q2 updates and 2026.11 items such as HANA Cloud calculation-view semantic onboarding, replication-flow scheduling, runtime-settings transport, SAP Snowflake, and SAPPHIRE announcementsPlatform Context:
references/business-data-cloud.md - SAP Business Data Cloud architecture, Databricks integration, data products vs. marketplace, BDC ConnectMCP Integration:
references/mcp-tools-reference.md - Complete MCP tool reference, 45 tools across 8 categories, API documentation, authentication patternsreferences/mcp-use-cases.md - 8 illustrative source-material use cases with personas and planning assumptionsreferences/edge-cdp-control.md - Datasphere-specific add-on for the shared sap-browser-automation Edge/CDP and authentication layerSearch MCP references before loading them. Use rg -n "<space|view|connection|task|lineage|marketplace|governance|admin>" references/mcp-*.md to locate the exact tool or scenario, then open only that excerpt.
references/mcp-tools-reference.md for tool names, inputs, and direct tenant-operation boundaries.references/mcp-use-cases.md only when choosing a workflow pattern or reviewing illustrative impact assumptions; do not load the full use-case guide for ordinary tool lookup.This plugin includes 3 specialized agents, 5 slash commands, and validation hooks:
Agents (in agents/):
datasphere-modeler - Data Builder tasks, views, flows, analytic modelsdatasphere-integration-advisor - Connectivity, replication, data integrationdatasphere-admin-helper - Space management, security, monitoringCommands (in commands/):
/datasphere-space-template - Generate space configurations/datasphere-view-template - Generate view templates (graphical/SQL)/datasphere-connection-guide - Step-by-step connection setup/datasphere-cli - CLI command reference and examples/datasphere-mcp-tools - SAP Datasphere MCP tool reference and usage guidanceHooks (in hooks/):
This skill includes a connection recipe for the SAP Datasphere MCP Server (@mariodefe/sap-datasphere-mcp). The bundled MCP config uses the approved exact pin 1.5.2, governed by sap-dependency-security and validated by npm run validate:mcp-security. Since v1.3.0 the server defaults to a lean tool profile (a reduced diagnostic surface); set DATASPHERE_TOOL_PROFILE=full in the MCP env to expose the complete tool set. Treat MCP behavior and live tenant tool counts as pending until your harness loads the server and verifies the available tools.
| MCP detail | Value | |------------|-------| | Command | npx | | Args | -y @mariodefe/sap-datasphere-mcp@1.5.2 | | Required env | DATASPHERE_BASE_URL, DATASPHERE_CLIENT_ID, DATASPHERE_CLIENT_SECRET, DATASPHERE_TOKEN_URL | | Operation safety | Tenant read tools plus mutating/destructive tools; ask before create, update, delete, reset, deploy, publish, or trigger operations | | Fallback | Use Datasphere CLI/reference guidance and mark live checks pending |
The MCP server enables:
See /datasphere-mcp-tools command for complete tool list.
OAuth 2.0 Client Credentials with automatic token refresh.
Required environment variables:
DATASPHERE_BASE_URLDATASPHERE_CLIENT_IDDATASPHERE_CLIENT_SECRETDATASPHERE_TOKEN_URLPerformance depends on tenant size, network latency, payload size, permissions, and server version. Treat any example timings or batch limits in MCP reference material as illustrative until measured in the target tenant and recorded as evidence.
plugins/sap-datasphere/
├── .claude-plugin/
│ └── plugin.json
├── .mcp.json # MCP server configuration
├── agents/
│ ├── datasphere-modeler.md
│ ├── datasphere-integration-advisor.md
│ └── datasphere-admin-helper.md
├── commands/
│ ├── datasphere-space-template.md
│ ├── datasphere-view-template.md
│ ├── datasphere-connection-guide.md
│ ├── datasphere-cli.md
│ └── datasphere-mcp-tools.md # MCP tools reference
├── hooks/
│ └── hooks.json
└── skills/
└── sap-datasphere/
├── .claude-plugin/
│ └── plugin.json
├── SKILL.md
├── README.md
└── references/
├── data-acquisition-preparation.md
├── data-modeling.md
├── graphical-sql-views.md
├── connectivity.md
├── administration.md
├── data-integration-monitor.md
├── data-access-security.md
├── content-transport.md
├── cli-commands.md
├── data-products-marketplace.md
├── catalog-governance.md
├── best-practices-patterns.md
├── whats-new-2025.md
├── whats-new-2026.md
├── business-data-cloud.md
├── mcp-tools-reference.md # MCP technical reference
├── edge-cdp-control.md # Edge CDP browser triage| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | fail→fail | 19,745 | 13,159 | -33% | 1 | 1 | 0% | 3,689 | 7,946 | +115% | 0 | 0 | — |
case-01 | fail→pass | 32,338 | 16,815 | -48% | 1 | 1 | 0% | 1,254 | 7,966 | +535% | 0 | 0 | — |
case-02 | fail→pass | 17,807 | 17,182 | -4% | 1 | 1 | 0% | 2,982 | 8,267 | +177% | 0 | 0 | — |
case-03 | pass→pass | 26,126 | 20,344 | -22% | 1 | 1 | 0% | 4,073 | 8,838 | +117% | 0 | 0 | — |
case-04 | pass→pass | 17,333 | 8,749 | -50% | 1 | 1 | 0% | 2,932 | 6,939 | +137% | 0 | 0 | — |
case-18 | pass→pass | 12,046 | 10,339 | -14% | 1 | 1 | 0% | 1,920 | 7,092 | +269% | 0 | 0 | — |
case-06 | pass→pass | 16,742 | 18,990 | +13% | 1 | 1 | 0% | 2,676 | 8,326 | +211% | 0 | 0 | — |
case-07 | fail→pass | 10,229 | 3,786 | -63% | 1 | 1 | 0% | 1,818 | 6,220 | +242% | 0 | 0 | — |
case-08 | fail→pass | 6,317 | 3,070 | -51% | 1 | 1 | 0% | 1,358 | 5,771 | +325% | 0 | 0 | — |
case-09 | pass→pass | 8,001 | 4,585 | -43% | 1 | 1 | 0% | 1,211 | 6,171 | +410% | 0 | 0 | — |
case-10 | pass→pass | 4,104 | 4,584 | +12% | 1 | 1 | 0% | 680 | 6,231 | +816% | 0 | 0 | — |
case-11 | pass→pass | 5,909 | 5,660 | -4% | 1 | 1 | 0% | 938 | 6,383 | +580% | 0 | 0 | — |
case-12 | pass→pass | 7,576 | 6,310 | -17% | 1 | 1 | 0% | 1,213 | 6,336 | +422% | 0 | 0 | — |
case-13 | pass→pass | 12,566 | 7,134 | -43% | 1 | 1 | 0% | 1,948 | 6,587 | +238% | 0 | 0 | — |
case-14 | pass→pass | 8,499 | 5,926 | -30% | 1 | 1 | 0% | 1,378 | 6,343 | +360% | 0 | 0 | — |
case-15 | pass→pass | 7,856 | 4,837 | -38% | 1 | 1 | 0% | 1,387 | 6,209 | +348% | 0 | 0 | — |
case-16 | pass→pass | 7,222 | 8,482 | +17% | 1 | 1 | 0% | 1,106 | 6,812 | +516% | 0 | 0 | — |
case-17 | pass→pass | 14,586 | 7,463 | -49% | 1 | 1 | 0% | 2,447 | 6,652 | +172% | 0 | 0 | — |
case-19 | pass→pass | 8,537 | 4,681 | -45% | 1 | 1 | 0% | 1,244 | 6,210 | +399% | 0 | 0 | — |
case-20 | pass→pass | 8,612 | 3,222 | -63% | 1 | 1 | 0% | 1,405 | 5,917 | +321% | 0 | 0 | — |
case-21 | pass→pass | 11,600 | 8,482 | -27% | 1 | 1 | 0% | 1,839 | 6,806 | +270% | 0 | 0 | — |
case-22 | fail→pass | 18,094 | 12,043 | -33% | 1 | 1 | 0% | 2,739 | 7,220 | +164% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +23 percentage points is the difference between those two pass rates over the 21 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.