Skip to content
Scalekit Docs
Talk to an Engineer Dashboard

Connect AI agents to Docparser

Scalekit connector
Open markdown

The Docparser connector lets your AI agent act in each user's Docparser account. Each user connects their own Docparser API key once, and Scalekit sends it with every call, so your agent never handles credentials. It comes with 10 tools.

Tools
10
What they doRead · write · destructive
6 · 4 · 06 read4 write0 destructive
Users sign in with

Setup

  1. Install the SDK

    Terminal window
    npm install @scalekit-sdk/node dotenv
  2. Set your credentials

    Add your Scalekit credentials to your .env file. Find values in app.scalekit.com > Developers > API Credentials.

    .env
    SCALEKIT_ENVIRONMENT_URL=<your-environment-url>
    SCALEKIT_CLIENT_ID=<your-client-id>
    SCALEKIT_CLIENT_SECRET=<your-client-secret>
  3. Create the Docparser connection

    In AgentKit > Connections, create a Docparser connection. The name you give it is the connection_name your code passes. See Configure connections.

Tools

Pass the exact name to execute_tool
Try in PlaygroundRequest a tool
  • docparser_document_result_getGet the parsed data of one document.Read-only

    Get Parsed Data For Document

    Get the parsed data of one document. Returns an array with one object holding document_id, remote_id, file_name, media links, page_count, uploaded_at, processed_at and every field the parser extracted. Use docparser_document_result_get when you know the document_id. Use docparser_results_list to get the data of many documents in a parser. Limited to 60 calls per minute. Requires a parser_id and a document_id from an import tool or docparser_results_list.

    Inputs

    document_idstringrequired
    ID of the document, as returned by docparser_document_fetch_url or docparser_document_upload_base64 (document_id or id) or by the document_id field of docparser_results_list. Example: 1d35639d4b53b59e77f737c93cd1d3d7
    parser_idstringrequired
    ID of the Document Parser, from docparser_parsers_list (the id field) or from Settings > API of the parser in the Docparser app. Example: mwekrupomwekrupo
    formatstring
    Shape of the parsed data. object (default) returns nested JSON objects; flat returns simple key/value pairs instead of nested objects. Example: flatone of objectflat
    include_childrenboolean
    Set true to also return the parsed data of child documents created during preprocessing, for example when a document was split. Example: true
  • docparser_document_status_getGet the processing state of one imported document.Read-only

    Get Document Status

    Get the processing state of one imported document. Returns file name, page count, in-progress flags for importing, processing and webhook dispatch, Unix timestamps for each stage (uploaded_at, imported_at, processed_at, etc.) and a failed_jobs list. Use docparser_document_status_get to check whether a document has finished parsing. Use docparser_document_result_get to read the parsed data once it has. Requires a parser_id and a document_id from an import tool.

    Inputs

    document_idstringrequired
    ID of the document, as returned by docparser_document_fetch_url or docparser_document_upload_base64 (document_id or id) or by the document_id field of docparser_results_list. Example: 1d35639d4b53b59e77f737c93cd1d3d7
    parser_idstringrequired
    ID of the Document Parser, from docparser_parsers_list (the id field) or from Settings > API of the parser in the Docparser app. Example: mwekrupomwekrupo
  • docparser_parser_models_listList the model layouts configured in one Document Parser.Read-only

    List Parser Model Layouts

    List the model layouts configured in one Document Parser. Returns an array of layouts, each with id and label. Use docparser_parser_models_list to see which document layouts a parser recognises. Use docparser_parsers_list to list the parsers themselves. Requires a parser_id from docparser_parsers_list.

    Inputs

    parser_idstringrequired
    ID of the Document Parser, from docparser_parsers_list (the id field) or from Settings > API of the parser in the Docparser app. Example: mwekrupomwekrupo
  • docparser_parsers_listList every Document Parser in the Docparser account.Read-only

    List Parsers

    List every Document Parser in the Docparser account. Returns an array of parsers, each with id and label. Use docparser_parsers_list to find the parser_id that every other Docparser tool needs. Use docparser_parser_models_list for the layouts inside one parser.

    Inputs

    This tool takes no inputs.

  • docparser_pingCheck that the Docparser API key on this connection is valid.Read-only

    Ping

    Check that the Docparser API key on this connection is valid. Returns {"msg": "pong"} when authentication succeeds. Use docparser_ping to test the connection. Use docparser_parsers_list to see what the account contains.

    Inputs

    This tool takes no inputs.

  • docparser_results_listList the parsed data of documents in one Document Parser, newest upload first by default, with filters by date, remote_id and sort order.Read-only

    List Parsed Data

    List the parsed data of documents in one Document Parser, newest upload first by default, with filters by date, remote_id and sort order. Returns an array of parsed-data objects, each with document_id, remote_id, file_name, media links, page_count, uploaded_at, processed_at and the extracted fields. Up to 100 documents by default. Use docparser_results_list to browse or filter results across a parser. Use docparser_document_result_get for a single known document. Limited to 30 calls per minute. Requires a parser_id from docparser_parsers_list.

    Inputs

    parser_idstringrequired
    ID of the Document Parser, from docparser_parsers_list (the id field) or from Settings > API of the parser in the Docparser app. Example: mwekrupomwekrupo
    datestring
    Cut-off date, required when list is uploaded_after or processed_after. An ISO 8601 date-time or a Unix timestamp. Example: 2024-02-12T15:19:21+00:00
    formatstring
    Shape of the parsed data. object (default) returns nested JSON objects; flat returns simple key/value pairs instead of nested objects. Example: flatone of objectflat
    include_processing_queueboolean
    Set true to also include documents that are not fully processed yet (still importing, preprocessing or parsing). By default only fully processed documents are returned. Example: true
    limitinteger
    How many documents to return when list is last_uploaded. Defaults to 100; at most 10000. Example: 50
    liststring
    Which documents to return. last_uploaded (default) returns the most recent uploads, newest first, paged with limit and offset. uploaded_after returns documents imported after date; processed_after returns documents parsed after date (both need date). Example: uploaded_afterone of last_uploadeduploaded_afterprocessed_after
    offsetinteger
    Number of documents to skip, for paging together with limit. Defaults to 0. Example: an offset of 20 with a limit of 10 starts at the 21st document: 20
    remote_idstring
    Return only documents imported with this remote_id (the ID from your own system set at import). Example: invoice-2024-0042
    sort_bystring
    Field to sort by. By default results follow upload order. Example: processed_atone of parsed_atprocessed_atuploaded_atfirst_processed_atimported_atintegrated_atdispatched_webhook_atpreprocessed_at
    sort_orderstring
    Sort direction: DESC (default, newest first) or ASC. Example: ASCone of ASCDESC
  • docparser_document_fetch_urlImport a document into a Document Parser by giving Docparser a publicly accessible URL to fetch it from.Write

    Import Document From URL

    Import a document into a Document Parser by giving Docparser a publicly accessible URL to fetch it from. Returns document_id, parser_id, remote_id and a message with the status URL; the document is fetched and parsed a few minutes later. Use docparser_document_fetch_url when the file is online. Use docparser_document_upload_base64 when you have the file content. Then poll docparser_document_status_get and read results with docparser_document_result_get. Requires a parser_id from docparser_parsers_list.

    Inputs

    parser_idstringrequired
    ID of the Document Parser, from docparser_parsers_list (the id field) or from Settings > API of the parser in the Docparser app. Example: mwekrupomwekrupo
    urlstringrequired
    Publicly accessible URL of the document (PDF, image, Word, etc.) for Docparser to download. It must be reachable without login. Example: https://www.example.com/invoices/inv-0042.pdf
    remote_idstring
    Optional ID from your own system to store with the document. Docparser keeps it through processing and returns it as remote_id in the parsed data, so you can match results to your records. Example: invoice-2024-0042
  • docparser_document_upload_base64Import a document into a Document Parser by sending its content as a base64 string.Write

    Import Document From Base64

    Import a document into a Document Parser by sending its content as a base64 string. Returns the new document id, file_size and account quota usage (quota_used, quota_left, quota_refill). Use docparser_document_upload_base64 when you have the file bytes. Use docparser_document_fetch_url when the file is at a public URL. Then poll docparser_document_status_get and read results with docparser_document_result_get. Requires a parser_id from docparser_parsers_list.

    Inputs

    file_contentstringrequired
    The whole file encoded as standard base64, without a data: URL prefix. Example: JVBERi0xLjQKJcfsj6IKNSAwIG9iago8PC9MZW5ndGggNiAwIFI+PgpzdHJlYW0K
    parser_idstringrequired
    ID of the Document Parser, from docparser_parsers_list (the id field) or from Settings > API of the parser in the Docparser app. Example: mwekrupomwekrupo
    file_namestring
    File name to store for the document, including its extension. If left empty, Docparser names the file after the upload time. Example: invoice-0042.pdf
    remote_idstring
    Optional ID from your own system to store with the document. Docparser keeps it through processing and returns it as remote_id in the parsed data, so you can match results to your records. Example: invoice-2024-0042