> ## Documentation Index > Fetch the complete documentation index at: https://docs.erdo.ai/llms.txt > Use this file to discover all available pages before exploring further. # Imports one batch of a file (a header and at most 1,000 rows) into an import > lead source. Every row is written as history: dated by the file's own submission time or default\_submitted\_at, never alerted on, billed or followed up until somebody starts it. A row already imported is not written again. dry\_run previews the batch. REST only, like the declaration. ## OpenAPI ````yaml /api/openapi.json post /v1/provider-lead-sources/import openapi: 3.0.0 info: title: Erdo API description: >- Erdo's REST API: query and write datasets, run agents, manage threads, integrations, pages, evals, workstreams, experiments, and bounded outreach. Authenticate with a Bearer API key (erdo_api_...) or scoped token (erdo_token_...). version: '2026-11-02' servers: - url: https://api.erdo.ai description: Production security: - bearerAuth: [] paths: /v1/provider-lead-sources/import: post: summary: > Imports one batch of a file (a header and at most 1,000 rows) into an import description: > lead source. Every row is written as history: dated by the file's own submission time or default\_submitted\_at, never alerted on, billed or followed up until somebody starts it. A row already imported is not written again. dry\_run previews the batch. REST only, like the declaration. operationId: POST:mcp.ImportLeadSourceRowsAPI requestBody: content: application/json: schema: properties: default_submitted_at: description: > of their own, when the mapping has none or a row's is blank. A file of contacts with no dates still needs one: history keeps each lead's own arrival, and Erdo never invents it. title: > default_submitted_at (RFC 3339) dates rows that carry no submission time type: string dry_run: title: > dry_run maps the rows and counts what would be written, writing nothing. type: boolean header: items: type: string title: | header is the file's header row. type: array name: title: | name is the import source's name. type: string rows: items: items: type: string type: array title: > rows are the file's rows, one string per cell, at most 1,000 a request. type: array type: object responses: '200': content: application/json: schema: properties: already_written: format: int64 title: > already_written is how many earlier imports or reads already wrote. type: integer created: format: int64 title: | created is how many leads new to the dataset were written. type: integer created_leads: description: > to start follow-up on. It includes leads an earlier request with the same rows created, so a batch sent again after a failure still names every one. items: type: string title: > created_leads are the permanent ids of the leads this batch's rows created, type: array existing: format: int64 title: > existing is how many matched a lead the dataset already has. type: integer no_contact: format: int64 title: | no_contact is how many have neither an email nor a phone. type: integer preview: $ref: '#/components/schemas/eventpipeline.LeadSourcePreview' refused: description: | more than one lead. format: int64 title: > refused is how many were held for review: contact details that point at type: integer unreadable: format: int64 title: > unreadable is how many the mapping cannot read (no id, no readable time). type: integer type: object description: Success response default: $ref: '#/components/responses/APIError' components: schemas: eventpipeline.LeadSourcePreview: properties: already_written: format: int64 title: | already_written is how many the ledger has already dealt with. type: integer existing_columns: description: | a surface can say a suggestion matches one. items: type: string title: > existing_columns are the lead dataset's own columns a source may write, so type: array format_columns: additionalProperties: type: string description: > one lead column each is read as when it is read: Meta's company\_name as company, campaign\_name as ad\_campaign\_name. A surface offers such a field only that column or nothing. title: > format_columns are the fields the format itself knows, by field, and the type: object header: description: | body's keys. items: type: string title: > header is the fields the source offers: a sheet's header row, a sample type: array history: description: > before the source was declared. No alert, no follow-up until started. format: int64 title: > history is how many would be written as history: leads who got in touch type: integer missing_fields: items: type: string title: | missing_fields are fields the mapping needs and the header lacks. type: array new: format: int64 title: | new is how many would be written as leads getting in touch now. type: integer not_read: items: type: string title: > not_read are the fields the mapping leaves out, which are never written. type: array read_columns: additionalProperties: type: string description: > column each becomes: what was named in columns and what the format reads by itself (Meta's company\_name as company), so a person sees every field a lead gets from the source. title: > read_columns are the fields the mapping reads, by field, and the lead type: object rows: format: int64 title: | rows is how many rows the source holds now. type: integer sample: items: type: object title: | sample is the first rows as they would be written. type: array skipped: additionalProperties: format: int64 type: integer title: | skipped counts rows that would not be written, by reason. type: object suggested_columns: additionalProperties: type: string description: | person to confirm. title: > suggested_columns proposes a lead column for each field not read, for a type: object suggested_values: additionalProperties: additionalProperties: type: string type: object description: > answers, the value each answer is stored as in its lead column, by field then answer: "are\_you\_a\_broker?" {"Yes, I'm an agent": "true"}. Saved as the mapping's values once a person confirms them. title: | suggested_values proposes, for a field holding a small fixed set of type: object test: format: int64 title: | test is how many the source marks as test submissions. type: integer title: | LeadSourcePreview is a dry run's account of what a source would write. type: object responses: APIError: content: application/json: schema: externalDocs: url: https://pkg.go.dev/encore.dev/beta/errs#Error properties: code: description: Error code example: not_found externalDocs: url: https://pkg.go.dev/encore.dev/beta/errs#ErrCode type: string details: description: Error details type: object message: description: Error message type: string title: APIError type: object description: Error response securitySchemes: bearerAuth: type: http scheme: bearer description: An Erdo API key (erdo_api_...) or scoped token (erdo_token_...). ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.