convertBatchResultsToEdgehancerResult({ batch, batchFetchIds: { invalidBatchIndices, validIdToOriginalIndexMap }, batchResults, resolveIdFromBatchResultFn, knownInvalidErrorMessage, missingBatchResultErrorMessage, }, input)

Given a batch of data resources and their batch-fetch results, convert them into a result that can be returned from a request hook.

getDataResourceAsRequest(data)

Converts a data resource to a Request object suitable to pass to fetch()

getDataResourceHeaders({ headers }, input)

Gets the HTTP headers for a data resource

getDataResourceQueryString({ parameters, url }, input)

Gets the query string for a data resource, returns a properly URL-encoded string

getDataResourceUrl(dataResource)

Gets the request URL for a data resource, including the url and any query parameters

resolveBatchFetchIds(batch, resolveBatchItemIdFn)

Given a batch of data resources to fetch, and a function to extract a batch-fetch ID from each resource, returns the valid IDs to fetch in a batch.

afterAIDataResourceEditContextSchema

afterAIDataResourceEditResultSchema

afterAIEditContextSchema

afterAIEditResultSchema

COLLECTION_DEFAULT_LIMIT

COLLECTION_DEFAULT_OFFSET

COLLECTION_MAX_LIMIT

createAIDataResourceEditContextSchema

createAIDataResourceEditResultSchema

createAIEditContextSchema

createAIEditResultSchema

dataResourceSchema

edgehancerMergedDataTypeSchema

mergedDataTypeSchema

preRequestEdgehancerDataResourceResultSchema

preRequestEdgehancerResultSchema

requestEdgehancerDataResourceResolutionResultSchema

requestEdgehancerResultSchema

resolvingIssueSchema

AfterAIDataResourceEditContext

AfterAIDataResourceEditHookFn

Shape of a hook function that is invoked after an AI edit has been performed on a data resource. The function can alter the success or value of the edit performed by the LLM.

This hook allows you to: - Apply custom business rules to AI-generated edits - Detect and correct hallucinations in AI output - Transform the AI result into a different format - Validate that the AI edit meets your specific requirements

NOTE: this function is only invoked when a value is edited. If the value is cleared or reset, the function is not invoked (those are internally handled).

NOTE: in most cases, you want createAIDataResourceEdit well before this hook. This hook allows catching hallucinations, applying deterministic business rules, or transforming the format of the AI result into another format.

AfterAIDataResourceEditResult

AfterAIEditContext

AfterAIEditHookFn

Shape of a hook function that is invoked after an AI edit has been performed to a custom property type. The function can alter the success or value of the edit performed by the LLM.

NOTE: this function is only invoked when a value is edited. If the value is cleared or overrides reset, the function is not invoked (those are internally handled)

NOTE: in most cases, you want createAIEdit well before this hook. This hook allows catching hallucinations, applying deterministic business rules, or transforming the format of the AI result into another format.

AfterAIEditResult

CollectedBinding

A normalized record of one bound parameter value that consumed a data resource somewhere in the composition tree. Custom edgehancer authors can use this to derive minimal upstream queries (e.g. GraphQL field selection, downstream select.* forwarding) instead of always fetching the full resource shape.

ConvertBatchResultsToEdgehancerResultOptions

CreateAIDataResourceEditContext

CreateAIDataResourceEditHookFn

Shape of a hook function that is invoked to create the prompt and output schema for an AI edit to a data resource.

This hook is called when AI needs to edit the content of a data resource. It should return instructions for the AI and a JSON schema describing the expected output format.

NOTE: the instructions returned by this hook are part of a larger system editing prompt. Only describe how to edit the data resource content itself, not general editing instructions.

CreateAIDataResourceEditResult

CreateAIEditContext

CreateAIEditHookFn

Shape of a hook function that is invoked to create the prompt and output schema for an AI edit to a custom property type.

NOTE: the prompt returned by this hook is part of a larger system editing prompt. Only describe how to edit the custom property type itself.

CreateAIEditResult

CustomEdgehancerDefinition

Declaration of custom edgehancer hooks on a merged data type

DataResource

A data resource of JSON data retrieved from a data type/data source

DataResourceFetchContext

The context we are fetching the data resource in:

• testing: when a data type is being tested in the UI using an ephemeral data resource - editing: the data resource is being fetched as part of editing a draft of a composition or entry - normal: the data resource is being fetched as part of rendering a composition or entry

NOTE: using the getDataResource() function from integration UI will have the same fetch context as the parent composition/entry data, i.e. 'editing' NOTE: this is subject to additive change

EdgehancerDataIssue

EdgehancerMergedDataType

MergedDataType

The result of merging a data source with a data type into a single request object

PreRequestEdgehancerContext

PreRequestEdgehancerDataResourceContext

PreRequestEdgehancerDataResourceResult

PreRequestEdgehancerResult

PreRequestHookFn

RequestEdgehancerContext

RequestEdgehancerDataResourceContext

RequestEdgehancerDataResourceResolutionResult

RequestEdgehancerResult

RequestHookFn

ResolveBatchIdsResult