Manage Classification Categories Setup Task Guidance

Manage Classification Categories Setup Task Guidance

Purpose and Boundary

Use this skill only to export or import the complete FSM setup package for Manage Classification Categories. Fix the task identity to HZ_MANAGE_CLASSIFICATION_CATEGORIES under Trading Community Foundation.

Do not use this skill for planning, sequencing, standalone validation, template generation, filtered exports, or individual classification-category inspection or maintenance. Validation is permitted only as a mandatory preflight for an export or import.

Supported Operations

1. export: export every setup-data object owned by the fixed task as one unfiltered, non-template FSM package through shared.setup-export.task-level-export.
2. import: import one complete, unchanged FSM ZIP for the fixed task through the mandatory core.data-loader workflow.

Treat a migration request as these two operations in sequence. Finish and persist the source export before starting a separate target import preflight and approval.

Single-Approval User Experience

For each export or import operation, complete all safe, read-only preparation before asking the user to approve anything. Intake questions, authentication, validation, preview generation, and corrections are preparation steps, not approval gates.

When preparation passes, present one compact execution summary containing the environment, operation, fixed task name/code, exact scope, artifact path and SHA-256 when applicable, output path, and side effect. Ask the user to reply exactly approve. That one approval authorizes the one immutable operation shown in the summary:

• for export, one FSM export submission plus polling and result retrieval;
• for import, one upload and initiation of the validated FSM ZIP.

Pass that approval through the selected shared or core workflow. Do not ask again at a task-local, Task Hook, adapter, upload, or initiation boundary. Downstream approval records are internal representations of the same user approval, not new prompts.

If input is missing, validation fails, the environment is ambiguous, or an explicit exception is required, stop with a friendly correction or blocker. Do not label the correction as an approval request. Any change to the approved environment, task, scope, archive bytes, hash, output path, or execution options invalidates the approval and requires a new prepared summary.

Fixed Task Context

• Pillar: CX
• Offering: Customer Data Management
• Functional area: Trading Community Foundation
• Functional area code: ORA_HZ_CDM_FOUNDATION
• Task name: Manage Classification Categories
• Task code: HZ_MANAGE_CLASSIFICATION_CATEGORIES
• CSV export/import supported: Y
• Data Loader import mode: FSM

Never accept a task-code, offering, or functional-area override.

Whole-Task FSM Export

1. Require project_id and the named source Fusion environment. Complete authentication, environment identity, privilege, live exact-task lookup, CSVExportImportSupportedFlag=Y, destination-path, and overwrite checks as read-only preflight; fail closed on a mismatch or blocker.

2. Fix template_only=false and export_criteria=[]. Refuse every filter, business-object subset, and template/specification request.

3. Build the single execution summary with the source environment, task name/code, unfiltered whole-task scope, output directory, and one-process impact. Ask exactly once for approve; do not ask for approval during discovery, authentication, validation, or output preparation.

4. Invoke shared.setup-export.task-level-export exactly once with the fixed context. Use its shared FSM engine for submission, polling, ZIP or ProcessLog retrieval, and packaging; do not implement task-local REST. Run the approved command with unbuffered output:

   bashCopy code
   python3 -u shared/setup-export/task-level-export/scripts/export_setup_task_csv.py \
     --data \
     --task-code HZ_MANAGE_CLASSIFICATION_CATEGORIES \
     --offering-name "Customer Data Management" \
     --functional-area-name "Trading Community Foundation" \
     --project-id "{project_id}" \
     --output-dir project/{project_id}/exports/customer-data-management/trading-community-foundation/manage-classification-categories/ \
     --token-cache auto \
     --wait-seconds 300 \
     --poll-interval 10
   

   If the command runner yields a session or cell identifier, wait on that same process until it exits. The -u option exposes the shared exporter's Started ... process: <ProcessId> line without changing shared code. Capture that identifier in the task skill's run record. Empty initial output is not failure and must not trigger a replacement export.

5. Persist the ProcessId as soon as it is returned. On success, verify the ZIP, calculate SHA-256, store it in the declared export directory, register it, and update project state.

6. On timeout, preserve the same ProcessId, last status, and elapsed time. Resume only that process after a later request; never submit a replacement export automatically.

Whole-Task FSM Import Through Data Loader

1. Require project_id, the named target Fusion environment, and one explicit ZIP path. Complete authentication, target identity, privilege, route, and source-path checks as read-only preflight. Reject loose CSV files, multiple archives, and directory scans.

2. Verify that the file is a readable ZIP containing exactly one member whose basename is ASM_SETUP_CSV_METADATA.xml, and require its single ObjectShortName to equal HZ_MANAGE_CLASSIFICATION_CATEGORIES. Preserve the archive byte for byte and calculate its SHA-256.

3. Invoke core.data-loader with the fixed task context and dataloader_import_mode=FSM. Let Data Loader own source review, task validation, FSM adapter selection, immutable preview creation, upload/initiation, and user-friendly reporting. Run its non-mutating preflight first:

   bashCopy code
   python3 core/data-loader/scripts/upload_and_initiate.py "{setup_data_zip_path}"
   

   Complete Task Hook validation and build the task-routed immutable preview before prompting the user. Require the preview and later approval envelope to retain executionRoute=data-loader, sourceType=FSM, the exact task code/name, absolute archive path, archive SHA-256, and wholeTask=true.

4. After every preflight check passes, display one user-friendly execution summary with the exact target, fixed task name/code, ZIP path and hash, whole-task scope, validation status, and configuration-changing impact. Ask exactly once for the separate message approve; stop before the first POST. On that message, record Task Hook approval for the displayed preview, verify the archive hash again, create the approval envelope, and immediately use Data Loader's approved task-routed command without another prompt:

   bashCopy code
   python3 core/data-loader/scripts/upload_and_initiate.py "{setup_data_zip_path}" \
     --task-routed \
     --task-validation-envelope "{task_validation_envelope_path}" \
     --approve-upload
   

5. Do not invoke setupTaskCSVImports, a shared import script, UCM upload, or a product adapter directly from this skill. Do not use an approval prompt to waive a validation problem: resolve it or report a blocker before the single execution summary. Any changed target, archive bytes, task context, or execution option invalidates approval and restarts Data Loader preflight.

6. Persist the accepted execution result and available redacted diagnostics in the declared report path. Check status only when the user asks, and never resubmit an import while its outcome is uncertain.

Error and Retry Contract

Classify failures as input_validation, catalog_or_metadata_mismatch, auth_expired_or_invalid, authorization_or_privilege, transport_or_environment, fusion_api_error, process_timeout, process_failed, artifact_persistence, or approval_blocker.

• Refresh a cache-managed token once after a read-only 401 or 403. A second authentication response blocks execution.
• Never retry an export or import POST automatically. Require a new immutable preview and explicit approval before any new submission.
• On timeout or failure, report the environment, operation, task code, ProcessId or request ID when available, last status, elapsed time, diagnostic artifact, and recommended next action.
• Never print or persist tokens, secrets, authorization headers, token-cache contents, full Base64 payloads, or sensitive customer data.

Result Contract

For export, report the source environment, fixed task name/code, ProcessId, status, ZIP or log path, enclosure type, and ZIP SHA-256 when available. For import, return the complete core.data-loader Markdown result and report the target environment, fixed task name/code, immutable ZIP path and hash, status, and available process log. Record the skill run and token usage for every durable project activity.

Refusal Rules

• Refuse individual classification-category creation, editing, deletion, or inspection.
• Refuse filtered, partial, criteria-based, and template/specification exports.
• Refuse task-code overrides, loose CSV files, edited or repacked ZIPs, and packages that Data Loader cannot confirm as the fixed FSM task.
• Refuse direct endpoint, script, or adapter execution that bypasses the required approval workflow.

Golden Scenarios

Sources

• knowledge/setup-task-catalog/effective/setupTasksAll.json
• shared/setup-export/task-level-export/SKILL.md
• core/data-loader/SKILL.md