{"ref":"artifact:sap-samples-cap-agentic-engineered","page":"/artifacts/sap-samples-cap-agentic-engineered","json":"/artifacts/sap-samples-cap-agentic-engineered.json","generated":"2026-10-26","drafted_with_ai":true,"entity":{"id":"sap-samples-cap-agentic-engineered","name":"SAP-samples/cap-agentic-engineered","artifact_type":"sample","ecosystem":"cap","flavors":[{"key":"node","source":{"url":"https://github.com/SAP-samples/cap-agentic-engineered","title":"SAP-samples/cap-agentic-engineered — README","checked":"2026-09-13","quote":"- [Node.js](https://nodejs.org/) >= 18"}}],"use_as":"starting-point","domain":"ai-and-agents","origin_ring":"sap-official","description":{"one_line":"A Financial Risk Analyzer built to show what changes when a coding agent is grounded by SAP's MCP servers rather than left to its training data — and it ships the AGENTS.md and skills that did the grounding.","short":"Financial Risk Analyzer built by an agent grounded in SAP MCP servers","what_it_solves":"The claim that agents write better SAP code when pointed at live documentation is usually asserted. This is the argument made as an application: a CAP service classifying general-ledger transactions as normal or anomalous over twenty-four engineered features, with an OData V4 service, a Fiori Elements UI and SAP AI Core behind it. What travels beyond the demo is the grounding itself — a reusable AGENTS.md carrying MCP routing rules, and a `skills/` folder that triggers on file patterns, routing the agent to the CAP server for models and services, the Fiori server for annotations and the UI5 server for controllers.","to_build_on_it":"The AGENTS.md and the skills are the portable part and are described as working across Claude Code, Cursor, Copilot and other MCP-capable tools; the application is the evidence they work. Taking the routing rules into a project of your own is a smaller step than taking the risk analyzer."},"links":{"repo":"https://github.com/SAP-samples/cap-agentic-engineered"},"type_properties":{"form":{"value":"reference-application","sources":[{"url":"https://github.com/SAP-samples/cap-agentic-engineered","title":"SAP-samples/cap-agentic-engineered — README","checked":"2026-10-18","quote":"A reference application showing how AI coding agents produce higher-quality SAP applications when grounded by"}]}},"identity":{"license":"Apache-2.0"},"skills":{"provenance":"indexed","kind":"capability","canonical_source":"https://github.com/SAP-samples/cap-agentic-engineered"},"demonstrates":["capability:agent-guidance","capability:reference-apps","capability:llm-integration"],"provenance":{"discovered_via":[{"source":"source:sap-ai-mcp-servers"}],"first_seen":"2026-09-09"},"history":[{"date":"2026-09-09","field":"entry","from":null,"to":"created","reason":"Entered from the sap-ai-mcp-servers catalogue, which is credited as the source; written from the repository's own README rather than from the catalogue's line about it. Hard rule 7: skills are indexed, never authored here. The catalogue files it under skills; its own README calls it a reference application showing what grounding does, so it is typed sample and demonstrates rather than provides. VISION §5: evidence is never an option you depend on."}],"drafted_with_ai":true,"maintainers":["actor:sap-samples"],"contributors":["actor:guilherme-segantini"]},"effective_normative_relations":[],"readings":{"liveness":[{"label":"Maintenance","tier":"unknown","evidence":"no release data"},{"label":"Continuity","tier":"good","evidence":"3 contributors, org-owned"},{"label":"Repository","tier":"good","evidence":"pushed 2026-05, not archived"}],"compatibility":[{"label":"Declared for","tier":"fair","evidence":"nothing declared"},{"label":"Survived majors","tier":"unknown","evidence":"no release dates"},{"label":"Maintainer's CI","tier":"fair","evidence":"no CI configured"}],"practice":[{"label":"Tests","tier":"good","evidence":"49 test files"},{"label":"Docs","tier":"good","evidence":"README 7 KB"},{"label":"Changelog","tier":"good","evidence":"CHANGELOG in the repo"}],"reach":[{"value":null,"unit":"installs/week"},{"value":14,"unit":"watching"}]},"signals":{"crawled_at":"2026-10-26","npm":"not published under this name","stars":14,"forks":3,"open_issues":1,"archived":false,"pushed_at":"2026-05-27","license_spdx":"Apache-2.0","org_owned":true,"default_branch":"main","contributors":3,"tree_truncated":false,"test_files":49,"workflows":0,"other_ci":false,"readme_bytes":7360,"docs_files":0,"changelog":true,"license_file":true,"dts_files":0,"readme_path":"README.md","readme_sha":"47bca57c0157c35bd1a8f87d28dc869a9dcdd909","created_at":"2026-04-14T17:07:26Z","open_prs":0,"first_commit":"2026-04-14","last_commit":"2026-05-27","commits":18,"repo_read":"2026-10-26"},"edges":{"maintainers":[{"ref":"actor:sap-samples","name":"SAP Samples","json":"/actors/sap-samples.json"}],"contributors":[{"ref":"actor:guilherme-segantini","name":"Guilherme Segantini","json":"/actors/guilherme-segantini.json"}],"explained_by":[],"alternative_to_this":[],"needed_by":[],"available_via":[],"parts":[],"add_ons":[],"answered_through_parts":[]}}