Manage Chart of Accounts Value Set Values
Purpose
Use this skill to create or update segment values in one named existing value
set. It prepares a values-only FBDI package and hands the verified package to
core/data-loader; Data Loader owns import selection, upload, submission,
monitoring, and returned outcome.
Use the parent Value Sets GET only to resolve the exact target and validate its
MaximumLength. Do not inspect requested values before or after the Data Loader
route.
Do not create or import hierarchies in this skill.
Input Contract
Accept four import sources:
| Source | Normalization |
|---|---|
| Natural-language text | Extract one ValueSetCode and every value row in input order |
| Image | Read visible table headers and cells, preserve blank versus populated fields, and ask once for unreadable required cells |
| CSV | Parse headerless FBDI rows with 13 through 101 columns |
| ZIP | Select exactly one GlSegmentValuesInterface.csv; a hierarchy CSV is not required |
Before authentication for an import, require:
- Value Set Code.
- At least one nonblank Value.
- Description for every value.
- Summary Flag for every value.
- Allow Posting Flag for every value.
- Allow Budgeting Flag for every value.
After flag normalization, reject any row where SummaryFlag=Y and either
AllowPostingFlag=Y or AllowBudgetingFlag=Y. A summary value must have both
Allow Posting and Allow Budgeting set to N. Never silently replace an
explicit conflicting flag; report the row and fields so the user can correct
the input.
A clear batch-wide flag applies to every row. Ask one consolidated question for
all missing required fields. Do not infer a Value or required flag from a
description or an unreadable image cell. A Description may support an Account
Type suggestion only under the reviewed Natural Account workflow below. Default
EnabledFlag=Y for text input when omitted; CSV and ZIP rows must supply
Enabled Flag in column 6.
Treat an attachment without an explicit create, import, upload, or execute
instruction as preparation/preview only. An explicit import instruction requests
live execution but does not approve a generated package. Route the immutable
package through core.data-loader, show its exact preview, and obtain explicit
approval before upload or submission. Any source, mapping, validation, or package
change invalidates that approval and requires a new preview.
Read-Only Value Set Preflight
After base input validation and authentication, but before package generation or handoff, resolve the target through the parent Value Sets collection:
GET /fscmRestApi/resources/11.13.18.05/valueSets
?q=ValueSetCode='{ValueSetCode}'
&fields=ValueSetId,ValueSetCode,MaximumLength,ValidationType,ValueDataType,ValueSubtype,ProtectedFlag
&limit=2
URL-encode the query and safely escape the supplied code. Apply an exact,
case-sensitive comparison to returned ValueSetCode values even when the
server-side query is broader.
- No exact match: report that the Value Set Code doesn't exist and stop without child-value retrieval, ZIP creation, upload, or ESS submission.
- Exactly one exact match: retain its
ValueSetId,ValueSetCode,MaximumLength, validation/data/subtype metadata, and protected state. - More than one exact match: stop as an ambiguous catalog result.
- Missing or nonnumeric
MaximumLength, or a value outside 1 through 150: stop as a catalog mismatch; never guess or default the length.
Use this parent GET only for existence and read-only metadata. Never use parent POST/PATCH/DELETE or child Values REST mutation in this skill. If the lookup is unauthorized or unavailable, report the auth, authorization, transport, or Fusion API failure and stop.
Value-Existence Handling
Never query child values or inspect existing rows before submission. This applies equally to Natural Account and normal segments. Let the Data Loader route determine whether supplied rows create, update, fail validation, or produce zero new values.
Do not offer or execute a standalone value-existence lookup from this skill. If a user asks whether values exist, explain that this task intentionally relies on the Data Loader result for create/update processing.
Determine value-set usage without querying values:
- In an end-to-end plan, use the established Chart of Accounts Structure Instance context when it proves that the target value set is assigned to the Natural Account segment.
- For a standalone task run, treat Natural Account usage as true only when the user explicitly declares it. Do not infer it from the value set name, value, or description.
- When Natural Account usage is proved or declared, require Account Type for every row. Suggest a type from the Description only when the meaning is unambiguous, then require the user to accept or correct every suggestion. Ask the user for any row that cannot be safely suggested.
- When Natural Account usage is not proved or declared, leave Account Type blank unless the user explicitly supplies it; validate and normalize any supplied value.
FBDI CSV Contract
Generate exactly 101 columns without a header. The first 13 positions are:
| Position | Field |
|---|---|
| 1 | Value Set Code |
| 2 | Value |
| 3 | Start Date |
| 4 | End Date |
| 5 | Summary Flag |
| 6 | Enabled Flag |
| 7 | Account Type |
| 8 | Allow Posting Flag |
| 9 | Allow Budgeting Flag |
| 10 | Third Party Control Account |
| 11 | Reconcile Flag |
| 12 | Financial Category |
| 13 | Description US |
Positions 14 through 101 retain the current template order for translated
descriptions, Attribute1 through Attribute50, Data Source, the four added
translated descriptions, and the workbook-only Is Validated marker. Always
write the Is Validated position blank. Use UTF-8 without a byte-order mark.
Follow the exact ordered mapping and deterministic processing algorithm in
ora-resource/ora_fbdi_processing.md.
Package the CSV as:
glsegmentvaluesandhierarchy.zip
└── GlSegmentValuesInterface.csv
Do not add GlSegmentHierInterface.csv for a values-only request. The 26C
template macro skips an empty hierarchy sheet, and the attached working sample
contains only the values CSV.
The loader control file supplies STATUS_CODE=NEW, OBJECT_VERSION_NUMBER=1,
LOAD_REQUEST_ID, and audit columns. Do not add those as CSV positions.
Validation and Defaults
Run complete-batch local validation before ZIP generation:
- Reject blank and duplicate values, and reject a blank Description.
- Enforce
MaximumLengthreturned by the exact parent Value Sets lookup. - Normalize Y/Yes/true/1 and N/No/false/0 to Y/N.
- Accept input dates as YYYY-MM-DD or YYYY/MM/DD; write YYYY/MM/DD and require End Date not earlier than Start Date.
- Apply the reserved-word, restricted-character, and description rules in
ora-resource/ora_input_validation.md. - Preserve supported translated descriptions, Attribute1 through Attribute50, and Data Source from CSV/ZIP input.
- Reject unknown nonblank columns after position 101.
For a value set proved or declared to be Natural Account, require a reviewed Account Type for every row. Normalize Asset/A, Liability/L, Expense/E, Owner's Equity/O, and Revenue/R to A/L/E/O/R, then default omitted Third Party Control Account and Reconcile Flag to N. For any other value set, preserve an explicit valid Account Type when supplied; otherwise leave it blank.
Do not create a package when any row fails.
Data Loader Handoff
After successful validation and package verification, produce a preview with:
- Exact environment and Value Set Code.
- Every requested value and business attribute.
- Every applied default.
- Row count, 101-column count, ZIP member list, CSV hash, and ZIP hash.
- Data Loader route and wait configuration.
Use this component split:
| Folder | Role |
|---|---|
core/data-loader/ |
Mandatory orchestration, validation, immutable preview, and approval workflow |
core/data-loader/ |
Validate task scope, construct the immutable preview, obtain approval, select the import route, and return the execution outcome |
Route every live create/update through core/data-loader. It owns the immutable
preview, approval, import selection, submission, monitoring, and returned
outcome. This task supplies only the verified values-only FBDI package and the
task-specific validation evidence. Do not add a second transport, REST, or ESS
execution path in this skill.
Treat the Data Loader result as the overall outcome. Report requested values as submitted and mark row-level created, updated, rejected, and unchanged counts as unavailable unless the returned result supplies them. Never run a post-handoff value lookup to manufacture those counts.
User-Facing Response
For create/update operations, return only the Data Loader outcome, Value Set Code, requested values, primary request ID when returned, terminal status, and concise loader diagnostics when applicable. State that row-level create/update/reject counts are unavailable from this route. Don't include documentation links, API links, local artifact links or paths, package or UCM document IDs, or internal storage details. Continue to persist prescribed audit artifacts internally.
Workflow
- Validate the create/update candidate from text, image, CSV, or ZIP input.
- Resolve exactly one parent value set by read-only GET and retain its
MaximumLength; stop on no match or ambiguity. - Determine Natural Account usage from established Structure Instance context or an explicit standalone declaration. For Natural Account values, suggest Account Type from Description where unambiguous and obtain user acceptance or correction; request any unresolved types.
- Generate or safely reuse and reopen the 101-column values-only ZIP and persist the preview and hashes.
- Hand the verified immutable package to
core/data-loaderfor its preview, approval, import selection, submission, monitoring, and outcome handling. - Do not issue separate REST, ESS, process-tree, child-job, LOG, OUT, or post-handoff value requests.
Markdown Processing Contract
This skill contains no local executable code. Read
ora-resource/ora_fbdi_processing.md before parsing a file, classifying segment
usage, validating rows, or constructing an FBDI package. Apply its algorithms
directly with the runtime's available file and archive capabilities. Never
continue to package generation without a successful classified prepared
candidate.
Read Order
ora-resource/ora_input_intake.mdora-resource/ora_input_validation.mdora-resource/ora_dependency_readiness.mdora-resource/ora_reference_lookups.mdora-resource/ora_defaulting_rules.mdora-resource/ora_bulk_intake.mdora-resource/ora_fbdi_processing.mdora-resource/ora_design_enforcement.mdora-resource/ora_error_handling.md
Boundaries
- Use the FBDI Data Loader route only for value creation and update.
- Use only a read-only parent Value Sets GET to prove exact existence and obtain
MaximumLengthand basic definition metadata. - Do not use child Values REST. Do not use Value Sets parent mutation.
- Do not query whether supplied values already exist, before or after import.
- Do not query values to classify Natural Account usage; use established Structure Instance context or an explicit standalone declaration.
- Do not run a post-handoff value or separate status call; determine the outcome
only from
core.data-loader. - Do not use FSM import/export, task-level setup export, or a fresh baseline export anywhere in this task.
- Do not create hierarchy rows or include a hierarchy CSV.
- Do not delete values.
- Route every create/update through
core.data-loader. - Do not expose credentials, tokens, UCM security details, or authorization headers.
Oracle Sources
- https://docs.oracle.com/en/cloud/saas/financials/26c/faigl/how-segment-value-and-hierarchy-import-data-is-processed.html
- https://docs.oracle.com/en/cloud/saas/financials/26c/oefbf/index.html
- https://docs.oracle.com/en/cloud/saas/applications-common/26b/farca/op-servername-fscmrestapi-resources-11.13.18.05-valuesets-get.html