Data and mappings

Access boundary: Current capability and authority required.

Implemented; private-source and optional AI-transfer gates apply.

Surfaces: Web, REST and MCP; private source/mapping ownership.

Read the engineering guide · All features

Capabilities

1. CSV, XLSX, XML and JSON imports with a private original
2. Sheet/header selection and XML record paths
3. English generated Data sheet names, with provenance-bound legacy aliases; uploaded XLSX and explicitly named JSON tables require exact mapping and join names
4. Profiling, anomalies and paginated inspection
5. Resume analysis of the same original
6. Versioned field/group/join mappings
7. Mapping validation before generation
8. Optional AI analysis under browser-administrator policy and transfer consent
9. Document-studio exclusions also cover datasets, direct records and AI functions; stop reading/profiling/mapping/generation on known excluded content without collecting its values

REST operations

• GET /api/datasets
• POST /api/datasets
• GET /api/datasets/{id}
• POST /api/datasets/{id}/retry-analysis
• GET /api/datasets/{id}/profile
• POST /api/datasets/{id}/analyze
• GET /api/mappings
• POST /api/mappings
• GET /api/mappings/{id}
• POST /api/mappings/{id}/validate

Read the REST integration guide for authentication, errors and contract discovery.

MCP tools

• analyze_dataset
• create_mapping
• get_dataset
• get_dataset_ai_policy
• get_mapping
• import_dataset
• import_json_dataset
• list_datasets
• list_mappings
• profile_dataset
• retry_dataset_analysis
• validate_mapping

Read the MCP integration guide for scope, approval and host boundaries.