mcpField
H
harmonizing-datacloud

harmonizing-datacloud

github.com/forcedotcomData & databases1.1k2.5k installsUpdated 1d ago
Repository

About harmonizing-datacloud

A skill for Salesforce Data Cloud harmonization, including data model objects, mappings, identity resolution, unified profiles, and data graphs. It guides schema inspection and controlled harmonization workflows for Data Cloud practitioners.

Key features

  1. 01Data model object inspection
  2. 02Field mapping workflows
  3. 03Identity resolution rulesets
  4. 04Unified profile workflows
  5. 05Data graph management

Use cases

  1. 01Reviewing source and target schemas
  2. 02Mapping source data to canonical models
  3. 03Running identity resolution

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/forcedotcom/sf-skills --skill harmonizing-datacloud

After installing, type/harmonizing-datacloudin Claude Code to confirm it loaded.

Bundled filesCREDITS.mdREADME.md

Details

From the project README

harmonizing-datacloud: Data Cloud Harmonize Phase

Use this skill when the user needs schema harmonization and unification work: DMOs, field mappings, relationships, identity resolution, unified profiles, data graphs, or universal ID lookup.

When This Skill Owns the Task

Use harmonizing-datacloud when the work involves:

  • sf data360 dmo *
  • sf data360 identity-resolution *
  • sf data360 data-graph *
  • sf data360 profile *
  • sf data360 universal-id lookup

Delegate elsewhere when the user is:

Required Context to Gather First

Ask for or infer:

  • source DLO and target DMO names
  • whether the task is schema creation, mapping, IR, or graph-related
  • target org alias
  • whether a ruleset already exists
  • the user’s desired unified entity model

Core Operating Rules

  • Inspect DMO schema before creating mappings.
  • Run the shared readiness classifier before mutating harmonization assets: node ../orchestrating-datacloud/scripts/diagnose-org.mjs -o <org> --phase harmonize --json.
  • Prefer dmo list --all when browsing the catalog, but use first-page dmo list for fast readiness checks.
  • Use query describe or dmo get --json instead of inventing unsupported describe flows.
  • Treat identity resolution runs as asynchronous and verify results after execution.
  • Keep unified-profile work separate from STDM/session tracing work.

Recommended Workflow

1. Classify readiness for harmonize work
node ../orchestrating-datacloud/scripts/diagnose-org.mjs -o <org> --phase harmonize --json
2. Inspect the catalog
sf data360 dmo list --all -o <org> 2>/dev/null
sf data360 identity-resolution list -o <org> 2>/dev/null
3. Inspect schema before mapping
sf data360 query describe -o <org> --table ssot__Individual__dlm 2>/dev/null
sf data360 dmo get -o <org> --name ssot__Individual__dlm --json 2>/dev/null
4. Create or review mappings intentionally
sf data360 dmo mapping-list -o <org> --source Contact_Home__dll --target ssot__Individual__dlm 2>/dev/null
sf data360 dmo map-to-canonical -o <org> --dlo Contact_Home__dll --dmo ssot__Individual__dlm --dry-run 2>/dev/null
5. Run IR only after mappings are trustworthy
sf data360 identity-resolution create -o <org> -f ir-ruleset.json 2>/dev/null
sf data360 identity-resolution run -o <org> --name Main 2>/dev/null

High-Signal Gotchas

  • dmo list should usually use --all.
  • Use query describe or dmo get --json; there is no dmo describe command.
  • Mapping and related commands can be sensitive to API-version differences.
  • Unified DMO names are ruleset-specific rather than generic.
  • Data graph definitions are sensitive to field selection and relationship shape.
  • If dmo list works but identity-resolution list is gated, treat that as a phase-specific gap rather than a full Data Cloud outage.

Output Format

Harmonize task: <dmo / mapping / relationship / ir / data-graph>
Source/target: <dlo → dmo or ruleset/graph names>
Target org: <alias>
Artifacts: <json files / commands>
Verification: <passed / partial / blocked>
Next step: <segment / retrieve / follow-up>

References

Comments

Markdown supported