CapAI Agent Hub
Enterprise teams need a governed way to configure, execute and observe AI agents around SAP work.
Private research prototypeOpen capabilityProduct ecosystem
Understand the relationships before opening the catalogue.
The flagship platforms organize evidence, governed intelligence and engineering control. Supporting capabilities extend testing, diagnostics, migration, performance and accountable decision-making.
Flagship relationship
Intelligence, engineering control and release evidence.
Start with the capabilities that explain how the wider ecosystem works.
CapAI Engineering Control Center
Agent platforms need engineering governance across releases, diagnostics, regression, migration and performance.
Private research prototypeOpen capability
CapAI Webpage Regression Bot
Website releases can introduce hidden functional, visual, navigation, flow, typography, responsive, browser, accessibility, SEO, security and performance regressions.
Build nearly complete — validation in progressOpen capabilityOutcomeEvidence-led human decision
Tools organize evidence and controls. Accountable professionals retain authority.
01SAP and enterprise systemsBusiness work, operational context and accountable system boundaries
02Evidence, APIs and adaptersOData, integrations, approved evidence and controlled data access
03Trust, identity and policySecurity, approvals, publication boundaries and human accountability
04Agent Hub and intelligenceGoverned agent catalogue, execution, trace and decision support
05Engineering Control CenterTesting, diagnostics, performance, migration and release control
06Human decisions and outcomesAccountable actions, recovery and measurable enterprise value
Complete product ecosystem
Every capability with its maturity and boundary.
Nothing has been removed. Open the detail only when the capability is relevant to your question.
01
Private research prototype
CapAI Agent Hub
Enterprise teams need a governed way to configure, execute and observe AI agents around SAP work.
- Target users
- SAP architects, platform owners, operations and governance teams
- Core capability
- Agent catalogue, system and service mapping, evidence setup, execution trace, workflow, telemetry, health, diagnostics, audit and controlled-action boundaries.
- Relationship to Agent Hub
- Primary governed intelligence and execution control plane.
- Relationship to ECC
- Validated and evolved through engineering controls, diagnostics, testing and release evidence.
- Publication boundary
- Architecture-level description only. No credentials, packages, source code or private builds are public.
- Source availability
- Not publicly available
02
Private research prototype
CapAI Engineering Control Center
Agent platforms need engineering governance across releases, diagnostics, regression, migration and performance.
- Target users
- Product architects, release managers, test and engineering teams
- Core capability
- Release governance, architecture decisions, build validation, regression, migration controls, diagnostics, performance analysis, test-data planning and quality gates.
- Relationship to Agent Hub
- Validates and governs the technical evolution of Agent Hub and connected components.
- Relationship to ECC
- This is the engineering companion and source of technical evidence.
- Publication boundary
- Concept and sanitised architecture only.
- Source availability
- Not publicly available
03
Build nearly complete — validation in progress
CapAI Webpage Regression Bot
Website releases can introduce hidden functional, visual, navigation, flow, typography, responsive, browser, accessibility, SEO, security and performance regressions.
- Target users
- Website owners, quality engineers, release managers and product engineering teams
- Core capability
- Baseline-led website discovery and regression assurance across routes, features, tabs, user flows, fonts, visual differences, responsive layouts, browser compatibility, accessibility, performance, APIs, security smoke checks, evidence and release gates.
- Relationship to Agent Hub
- Can use governed agent assistance for feature discovery, test-case suggestions and failure summaries, while deterministic test execution remains authoritative.
- Relationship to ECC
- Designed as a protected Regression Testing capability within the Engineering Control Center, with CI-based execution and an evidence-led release-gate view.
- Publication boundary
- Public product description only. Implementation details, baselines, screenshots, traces and validation evidence remain protected.
- Source availability
- Not publicly available
04
Private research prototype
CapAI Migration Control Center
Complex migrations need traceable readiness, dependency and cutover controls.
- Target users
- Migration leads, architects and release teams
- Core capability
- Migration planning, object readiness, dependency control and evidence-led gates.
- Relationship to Agent Hub
- Can expose governed migration intelligence as agent capabilities.
- Relationship to ECC
- Feeds migration validation and readiness evidence into ECC.
- Publication boundary
- Capability description only.
- Source availability
- Not publicly available
05
Private research prototype
CapAI Regression Testing
Fast product change increases regression risk across connected capabilities.
- Target users
- Test architects, release teams and product engineering
- Core capability
- Risk-based regression planning, traceability and result evidence.
- Relationship to Agent Hub
- Validates agent workflows and capability behaviour.
- Relationship to ECC
- Acts as the regression control component for ECC.
- Publication boundary
- Capability description only.
- Source availability
- Not publicly available
06
Private research prototype
CapAI Diagnostics
Failures across UI, APIs, runtime and integrations are difficult to isolate quickly.
- Target users
- Operations, developers and architects
- Core capability
- Cross-layer diagnostics, RCA support, telemetry correlation and recovery guidance.
- Relationship to Agent Hub
- Diagnoses agent execution, system mapping and runtime failures.
- Relationship to ECC
- Provides technical evidence for release and performance decisions.
- Publication boundary
- Capability description only.
- Source availability
- Not publicly available
07
Concept roadmap
CapAI Test Data Factory
Representative test data is difficult to plan without exposing sensitive production data.
- Target users
- Test, migration and engineering teams
- Core capability
- Policy-led test-data planning, generation patterns and coverage evidence.
- Relationship to Agent Hub
- Potential governed agent capability for safe test-data orchestration.
- Relationship to ECC
- Planned ECC component for test readiness.
- Publication boundary
- Concept only; no implementation claim.
- Source availability
- Not available
08
Independent research architecture
Enterprise Intelligence Adapter
Enterprise systems need a governed adapter between evidence, models, policies and operational decisions.
- Target users
- Enterprise architects and AI platform teams
- Core capability
- Evidence mediation, trust boundaries, provider abstraction and controlled intelligence patterns.
- Relationship to Agent Hub
- Extends governed evidence and intelligence interaction.
- Relationship to ECC
- Subject to architecture, security and performance gates.
- Publication boundary
- Independent research architecture; restricted implementation details are excluded.
- Source availability
- Selected public architecture material may be released separately
09
Independent research architecture
Enterprise Agent Fabric
Distributed enterprise agents require identity, policy, interoperability and lifecycle governance.
- Target users
- Enterprise and AI architects
- Core capability
- Agent identity, work routing, trust, evidence and human accountability patterns.
- Relationship to Agent Hub
- Agent Hub can operate as a governed control plane within the fabric.
- Relationship to ECC
- ECC governs engineering quality and release evolution.
- Publication boundary
- Architecture-level description only.
- Source availability
- Not publicly available
10
Technical proof validated with target-system and provider constraints
ABAP4C Cloud Deployment Pipeline — Architecture 1
ABAP Cloud delivery teams need a standards-led automation path that can move from an AI or coding agent through Eclipse ADT MCP tooling into SAP ADT execution while preserving audit evidence and clear proof boundaries.
- Target users
- SAP architects, ABAP Cloud developers, automation engineers and delivery teams evaluating standards-led ADT MCP automation
- Core capability
- AI or Coding Agent → VS Code / Cline / Python → Standard ADT MCP Server in Eclipse → SAP ADT tools → ABAP Cloud → audit and execution evidence.
- Relationship to Agent Hub
- Represents the direct standards-led agent-to-ADT MCP path for ABAP Cloud automation.
- Relationship to ECC
- Evidence is interpreted through release, validation and proof boundaries rather than broad capability claims.
- Publication boundary
- Architecture-level description with bounded live-evidence language.
- Source availability
- Selected public architecture material may be released separately
11
Independent engineering prototype with bounded live execution evidence
ABAP4C Cloud Deployment Pipeline — Architecture 2: Capella
ABAP Cloud deployment sometimes needs a dedicated control plane, local Eclipse bridge, bounded execution controls, receipts and recovery evidence rather than a direct tool invocation path alone.
- Target users
- SAP architects, platform owners, governed-delivery teams and automation engineers needing evidence-first ABAP Cloud deployment controls
- Core capability
- AI or Coding Agent → VS Code / Cline / Python → Capella Control Plane → Local Eclipse Bridge → ADT / RIS / DDIC execution → SAP ABAP Cloud → receipts, evidence and rollback controls.
- Relationship to Agent Hub
- Represents the governed-bridge route for ABAP Cloud deployment when policy, recovery and receipts matter.
- Relationship to ECC
- Capella-style execution is governed through explicit lifecycle, validation, approval, execution receipts and rollback references.
- Publication boundary
- Architecture-level description with bounded live-execution evidence language.
- Source availability
- Selected public architecture material may be released separately
12
Controlled-corpus validation completed; live-runtime coverage remains explicitly bounded
MCP Evaluation Framework
Teams need to prove whether an MCP-enabled tester detects seeded defects for the correct reasons rather than merely producing plausible reports or unsupported PASS outcomes.
- Target users
- SAP test architects, MCP developers, quality engineers and AI governance teams evaluating evidence-driven testing systems
- Core capability
- Adapter Framework, Seeded-Defect Discrimination Framework, clean-artifact baseline, defective-artifact corpus, PASS/FAIL/BLOCKED outcomes, false-PASS prevention, evidence normalisation and deterministic assertions.
- Relationship to Agent Hub
- Can expose testing capability as a governed agent-facing service while deterministic evidence remains authoritative.
- Relationship to ECC
- Provides measurable quality-gate evidence for SAP testing adapters and regression protection.
- Publication boundary
- Public product description and canonical publication relationship only.
- Source availability
- Selected public architecture material may be released separately