add_application_required_documents
ChatGPTAdds required documents (stips) to an application, taking their ids from list_document_types. Returns only the documents added, where list_application_required_documents lists them all.
check_permission
ChatGPTReports whether the authenticated user may perform up to 25 named abilities on named records, without performing any of them.
create_application
ChatGPTCreates an application from a workflow template, with Application-stage block data passed as additional top-level arguments keyed by block id, such as business or personal. list_workflow_templates lists the templates to choose from and get_workflow_template_sample gives one template's block ids and field names. Setting move_to_next_stage or with_validation makes the template's Application-stage required fields mandatory. move_to_next_stage can land the application in an underwriting stage, which starts a data orchestration run on its own, and an orchestration run, whether started that way or by data_orchestration_template_id, makes billable third-party bureau calls. A data_orchestration_template_id the requester is not permitted to run, or that cannot run against this application, is rejected. A submitted note saves even when notes comes back empty, which happens when the requester may not read the deal's notes.
create_application_note
ChatGPTAdds a note to an application, optionally against one workflow stage named by flow_id, which list_application_workflow_stages lists. It defaults to the stage the application currently sits on. Valid visibility scopes are listed by get_options. Each call adds another note rather than replacing one.
create_application_temporary_links
ChatGPTCreates expiring links a borrower opens to act on one application: sign consent, fill in personal information, connect a bank through Plaid or accounting software through Codat, verify identity through Vouched, view the stipulation files completed for an offer, or sign into the borrower platform. Set at least one flag. Owner IDs are listed under owners by get_application, and offer IDs by list_application_offers. get_application_links returns the DocuSign signing link, the resume link and the Rutter store connections. Nothing is sent to the borrower.
create_business_note
ChatGPTAdds a note to the business behind one application, where create_application_note keeps a note to the one application. A note that is not internal also lands on every application sharing that business. Each call adds another note rather than replacing one.
create_offer
ChatGPTCreates an offer on a placement of an application. The offer template's fields are passed as additional top-level arguments keyed by field name, and get_offer_template_sample lists them with the ones an offer requires, taking the template id from origin_offer_template in get_placement. stips takes document type ids from list_document_types. Each call adds another offer, and when the application is linked to another one through an internal funding client, it also copies the offer to that application and approves the matching placement there. update_offer changes an existing offer.
create_placement
ChatGPTCreates a lender placement on one application and submits it to the funder by method: Api queues a submission to the funder's integration and Email emails the funder. stage_id is a placement-type stage from list_application_workflow_stages, whose funder_ids lists the funders that stage accepts, and list_funders returns each funder's product_types and available_methods. A funder that belongs to an internal funding client also gets its own copy of the application, on the workflow template and offer template given in destination, each picked automatically when omitted. Each call adds another placement, except that Api refuses a matching earlier Api placement that has not failed and resubmits one that has. update_placement_status changes an existing placement.
create_task
ChatGPTCreates a task, such as a callback or follow-up, on one application, with a due date, a priority and optionally an assignee. A new task starts as todo, and update_task changes or completes it. get_current_user returns the requester's own ID for assignee_id, and list_assignable_users with ignore_assigned_deals set to true lists other users by role. Each call adds another task rather than replacing one.
duplicate_application
ChatGPTCreates a new application by copying an existing one onto a chosen workflow template, carrying over its business, owners, prequalification, Plaid data and tracking tokens, plus its documents unless sync_documents is false. list_workflow_templates returns the template ids. Offers, placements and notes are not copied, and create_application starts an application from block data instead. The copy starts the template's fast-track automation, which can move it into underwriting and make billable third-party bureau calls.
get_application
ChatGPTReturns one application by ID, where list_applications searches and filters across many. Offers, placements, documents and underwriting data are not included.
get_application_data_orchestration_log
ChatGPTReads one data orchestration run on an application step by step, where list_application_data_orchestration_logs returns one summary row per run, whose id is the orchestration_log_id here. The value each attribute check read is not returned. A run with no finished_at is still in progress, so its outcome is not final.
get_application_links
ChatGPTReturns borrower links for one application: a DocuSign signing link for an owner, a resume link back into the widget, a manual bank statement upload link, Codat and Rutter links to connect accounting software or an online store, and a Plaid Link token, which feeds an embedded Plaid SDK and is not a URL a borrower can open. Set a flag for each link wanted. With no flag set it returns the DocuSign link for the primary owner and a Plaid Link token. The first Codat request for a business, and the first Rutter request for each store platform, creates the connection with that provider. Owner IDs are listed under owners by get_application. create_application_temporary_links returns the expiring widget links for consent, personal information, Plaid, identity verification, completed stipulations and the borrower platform. Nothing is sent to the borrower.
get_application_workflow_sample
ChatGPTReturns the payload one existing application accepts, stage by stage: the block ids with an example value for every field in each block, and the field names each block requires. That block data is what update_application takes as top-level arguments. This reads the workflow snapshot the application is validated against, which is what update_application checks, where get_workflow_template_sample reads the template and so can differ from what an existing application accepts.
get_client
ChatGPTReturns one client organization by ID. list_clients returns id and name for many clients with search, filters and paging, and list_client_brand_names returns the brand names one client trades under.
get_communication_log
ChatGPTReads one communication log in detail. Takes the top-level id of a list_communication_logs row, not the id inside its type object.
get_current_user
ChatGPTReturns the authenticated Lendflow user's own profile, including their client ID, role names and assigned permissions preset. It carries no individual permissions, which check_permission answers for a named ability.
get_email_template
ChatGPTReturns one email template in full by ID, including the dynamic field tags it uses. list_email_templates finds templates by name, and list_dynamic_fields lists every tag a template can hold.
get_funder
ChatGPTReturns one funder by ID with its eligibility thresholds, restrictions and funding terms. list_funders returns the summary fields for many funders with search, filters and paging.
get_offer
ChatGPTReturns one offer by ID with its term and closing details, addressed without an application, where list_application_offers returns one thin row per offer on a given application.
get_offer_checkout_progress
ChatGPTReturns an accepted offer's checkout progress. Fails when checkout is not enabled in the application's workflow or has not started for the offer. list_application_offers returns each offer's status.
get_offer_funding_transaction
ChatGPTReads one funding transaction on an offer. list_offer_funding_transactions returns the offer's transactions with the same fields, and a row's id is the funding_transaction_id here. offer_id comes from list_application_offers.
get_offer_template_sample
ChatGPTReturns the fields one offer template takes, each with an example value, and which of them an offer and a closing require. The example values are generated placeholders, not defaults. get_offer and list_application_offers return an offer's offer_template, and get_placement a placement's offer_template and origin_offer_template. get_workflow_template_sample covers the application data a workflow template takes.
get_options
ChatGPTReturns the option and reference values behind Lendflow's tools and dashboard filters, resolving a name or a label into the ID another tool needs: product types, note visibility scopes, placement statuses and decline reasons, funder risk tiers and lender networks, and the industry, geographic, credit score and revenue lookup lists. Those lookup lists make the payload large. list_application_workflow_stages returns one application's stages.
get_placement
ChatGPTReturns one lender placement on an application in full, addressed by placement ID, where list_application_placements returns one thin row per placement.
get_workflow_template_sample
ChatGPTReturns the payload one workflow template accepts, stage by stage: the block ids with an example value for every field in each block, and the field names each block requires. That block data is what create_application and update_application take as top-level arguments, and list_workflow_templates returns the template ids.
kill_application
ChatGPTMarks one application as dead, recording a decline reason, sub-reason and notes. The reason and sub_reason also overwrite the workflow status and sub-status that update_application_workflow_status sets, and mark that status as no longer live. reason and sub_reason come from the statuses and sub-statuses list_application_workflow_statuses returns. Calling it on an application that is already dead replaces the recorded reason.
list_application_data_orchestration_logs
ChatGPTLists the data orchestration runs recorded on one application, newest first, where list_application_data_orchestration_templates returns the templates a stage can run. stage_id comes from list_application_workflow_stages. A run with no finished_at is still in progress, so its outcome is not final.
list_application_data_orchestration_templates
ChatGPTReturns the data orchestration templates one underwriting stage declares on an application's workflow snapshot. stage_id comes from list_application_workflow_stages and must name a stage of type Underwriting, so any other stage is rejected. A draft template is listed but cannot be run, which the status field distinguishes.
list_application_files
ChatGPTLists the files uploaded to one application, where list_application_required_documents lists the documents its workflow asks for and whether each is complete.
list_application_offers
ChatGPTReturns the offers on one application that have a funder attached, where get_offer returns one offer with its term and closing details.
list_application_placements
ChatGPTReturns the lender placements on one application, newest first, where get_placement returns one placement in full. Placements with no funder attached are excluded.
list_application_required_documents
ChatGPTLists the required documents for one application and whether each is complete, where list_application_files lists the files uploaded to it.
list_application_scorecard_groups
ChatGPTLists the finished scorecard groups available on an application, newest first, with optional search and client filters. A group combines several scorecards into one weighted score, and list_application_scorecards lists the individual scorecards. workflow_stage_id from list_application_workflow_stages narrows them to the groups that stage runs.
list_application_scorecards
ChatGPTLists the finished scorecards available on an application, newest first, with optional search, client and workflow template filters. workflow_stage_id from list_application_workflow_stages narrows them to the scorecards that stage runs. list_application_scorecard_groups lists groups of scorecards.
list_application_workflow_stages
ChatGPTReturns the stages defined in one application's workflow snapshot, in the order the application's flow runs them, with no branching information. Which stage the application currently sits on comes from get_application.
list_application_workflow_statuses
ChatGPTReturns the statuses and sub-statuses one application can be moved to, each with its name, color and visibility, as defined by the workflow template behind it. update_application_workflow_status takes a status and sub_status from this list. These are status values, not the workflow stages list_application_workflow_stages returns.
list_applications
ChatGPTSearches applications, with a large set of filters over people, dates, amounts, workflow status, offers, placements, score cards and tasks. get_options lists the values many of those filters accept, and list_funders the funder IDs the offer and placement filters take. Each row is a summary carrying the application name, client, workflow stage, amount and offer, placement and call counts, where get_application returns one application in full.
list_assignable_users
ChatGPTLists the users who can be assigned to an application in one role, such as underwriter or client analyst, ordered by last name. By default it also returns users still assigned to some application in that role, including deleted users and users who no longer hold it, and ignore_assigned_deals set to true leaves only current holders of the role. For a Lendflow user, a client role without client_id spans every client, so client_id narrows it to the application's own client. update_application_assignments assigns them.
list_call_summaries
ChatGPTReturns the summaries of one call, where list_call_transcriptions returns its transcription rows. Takes the id inside the call object of a list_communication_logs row as call_session_id, not the top-level id of that row.
list_call_transcriptions
ChatGPTReturns the transcription rows of one call, where list_call_summaries returns its summaries. Takes the id inside the call object of a list_communication_logs row as call_session_id, not the top-level id of that row.
list_client_brand_names
ChatGPTReturns the brand names configured on one client. get_client returns that client in full but carries none of them. create_application takes one of these ids as brand_name_id, and the brand_names filter on list_applications matches the names.
list_clients
ChatGPTReturns the id and name of each client organization, ordered by name, with optional client-name search and approved and is_funder filters. get_client returns one client in full by ID.
list_communication_logs
ChatGPTLists the calls, tasks, emails, SMS messages, notes and AI agent sessions recorded across Lendflow, including the notes that create_application_note and create_business_note add. Each row is a summary, and get_communication_log reads one log in detail. A call row only says whether the call has a transcription or a summary, and list_call_transcriptions and list_call_summaries return them.
list_document_types
ChatGPTLists the document types (stips) that can be required from a borrower on an application or an offer. list_application_required_documents returns the documents one application already requires.
list_dynamic_fields
ChatGPTReturns every dynamic field tag a template can hold, with its name and the template types it is allowed in. A tag can take options after a double underscore, such as [link.resume_link__utmSource=spring], and still counts as that tag; when its options_in_url is true, options become query parameters of the generated URL. get_email_template returns the tags one template uses.
list_email_templates
ChatGPTReturns the email templates the user can see, newest first, without their content, drafts and unapproved templates included. approved_only keeps only approved templates, published or not. get_email_template returns one template in full by ID.
list_filters
ChatGPTReturns saved application searches (SmartViews). list_applications takes one of these ids as filter_id and returns that SmartView's applications, and takes keys of its data as cleared_filters to leave those criteria out.
list_funders
ChatGPTReturns a summary row per funder, ordered by full_name, with optional search, risk tier, product type, country, area, city, approval and lender network filters, and selected_ids to order chosen funders first. get_funder returns one funder in full by ID.
list_offer_funding_transactions
ChatGPTLists the funding transactions recorded on one offer, the draws on a line of credit (LOC) or the invoices on an ARLOC, where get_offer_funding_transaction reads one of them. offer_id comes from list_application_offers.
list_workflow_templates
ChatGPTReturns the workflow templates, ordered by name, with nothing of each template's content. get_workflow_template_sample returns the payload a template accepts, and create_application accepts an active template that either the user's own organization owns or Lendflow publishes to every organization.
read_documentation_page
ChatGPTReturns the full markdown of one public page on docs.lendflow.com. search_documentation finds pages and returns the path this takes.
revive_application
ChatGPTRevives one declined application, restoring the status it held before it was declined and discarding the recorded decline reason, sub-reason and notes. An application that is not dead is rejected. The workflow status and sub-status still hold the decline reason afterwards, and update_application_workflow_status replaces them.
run_application_data_orchestration
ChatGPTRuns a data orchestration template on an application at the Underwriting stage named by stage_id, taking template_id from list_application_data_orchestration_templates for the same stage. A run can make billable third-party data pulls such as credit reports. The run is queued and returned as scheduled, and its id is the orchestration_log_id get_application_data_orchestration_log reads to follow its progress. The call fails while an unfinished run of the same template exists at that stage.
run_scorecard_groups
ChatGPTRuns scorecard groups on an application at the ScoreCards stage named by stage_id, which list_application_workflow_stages lists, adding a new result for each group. Without score_card_group_ids it runs every finished group that stage uses, which list_application_scorecard_groups with workflow_stage_id lists. The run is queued and returns an acknowledgement with no run identifier.
run_scorecards
ChatGPTRuns scorecards on an application at the ScoreCards stage named by stage_id, which list_application_workflow_stages lists, adding a new result for each scorecard. Without score_card_ids it runs every scorecard that parameter accepts, which can be more than the stage uses, and list_application_scorecards with workflow_stage_id lists the ones the stage uses. The run is queued and returns an acknowledgement with no run identifier.
search_documentation
ChatGPTSearches the public Lendflow API and product documentation on docs.lendflow.com for passages matching a query. read_documentation_page returns one page in full.
show_call_recording_player
ChatGPTRenders a player for the recording of one call, or for its voicemail when the call was missed. The audio reaches the user only, while list_call_transcriptions returns what was said.
show_upload_application_documents_app
ChatGPTRenders an interactive app the user can drop files into to upload documents to an application.
update_application
ChatGPTUpdates one application's Application-stage block data, tracking tokens, UTM parameters, client tracking token or brand, and moves it between workflow stages. Block data is passed as additional top-level arguments keyed by block id, such as business or personal, and get_application_workflow_sample gives the block ids and field names the application's own workflow snapshot accepts. Stage movement has three forms, where move_to_next_stage advances one stage, rollback returns to the previous stage, and move_to_stage_id jumps to a chosen stage by passing through every stage in between with force applied and no validation. The application must already have a workflow snapshot. Stage movement can start a data orchestration run, and an orchestration run, whether started that way or by data_orchestration_template_id, makes billable third-party bureau calls. A data_orchestration_template_id the requester is not permitted to run, or that cannot run against this application, is rejected. A submitted note saves even when notes comes back empty, which happens when the requester may not read the deal's notes.
update_application_assignments
ChatGPTAssigns or clears the Lendflow and client underwriter, analyst and funding advisor on one application, changing only the roles passed, where null clears one. Each assignee must be an active user holding the matching role, and for a client role a user of the application's own client, which list_assignable_users lists with ignore_assigned_deals set to true.
update_application_data_orchestration_log_master_status
ChatGPTSets the master status of one data orchestration run that is no longer running, named by orchestration_log_id from list_application_data_orchestration_logs, and records the optional comment in its history. The run does not resume; update_application_data_orchestration_log_step_outcome continues it instead.
update_application_data_orchestration_log_step_outcome
ChatGPTSets the outcome of the step one data orchestration run stopped at, named by orchestration_log_id from list_application_data_orchestration_logs, and resumes the run from there, or retries the step, which can repeat a billable third-party data pull. The action is queued, so the call only acknowledges it and get_application_data_orchestration_log shows the result. A run that is still running, a run already in the requested outcome, and a repeat of the same outcome while the earlier call is still queued all leave the run unchanged without an error. A different outcome queued behind an earlier one runs after it.
update_application_workflow_status
ChatGPTOverwrites the workflow status and sub-status on one application's workflow snapshot, leaving the application's own status column untouched. Omitting sub_status clears any existing sub-status, and every call stamps the status update time and resets the status color, so repeating identical values is still a write. Values are not checked against the workflow template's configured statuses that list_application_workflow_statuses lists, and an application with no workflow snapshot is rejected.
update_offer
ChatGPTChanges the terms of one existing offer on an application. The offer template's fields are passed as additional top-level arguments keyed by field name, and only the fields passed change. get_offer_template_sample lists the fields of the offer's template, whose id get_offer returns. stips takes document type ids from list_document_types. create_offer adds a new offer, and update_offer_sharing shows or hides one to the borrower. update_offer_status moves one to another status.
update_offer_required_documents
ChatGPTReplaces the required documents (stips) on one offer, such as a lender's conditions, with the list sent, where add_application_required_documents adds documents to the whole application. Ids come from list_document_types, and the current list from get_offer, whose required_count is sent back as count. update_offer changes the offer's terms and takes the same stips list alongside them. The borrower sees an offer's documents only once the offer is accepted.
update_offer_sharing
ChatGPTShows one existing offer to the borrower on the borrower platform, or hides it again, through is_shared. update_offer changes the offer's terms, and update_offer_status moves it to another status.
update_offer_status
ChatGPTMoves one existing offer to another status. Accepting or declining an offer that a funder sent through its integration can also send that decision to the funder, and when the application is linked to another one through an internal funding client, the linked copy of the offer moves too. update_offer changes the offer's terms, and update_offer_sharing shows or hides it to the borrower.
update_placement_status
ChatGPTSets the status of one lender placement belonging to a given application. It requires the ID of a placement-type stage from that application's current workflow, and a decline reason when the status is Declined. Declining also retracts that placement's offers, and no status change submits anything to the funder, while repeating the same values leaves the placement unchanged.
update_task
ChatGPTUpdates one task's description, due date, priority, assignee or status, leaving unsupplied fields unchanged. Setting status to done completes the task. task_id is the id inside a task entry from list_communication_logs, or the id create_task returns, not the communication log's own id. get_current_user returns the requester's own ID for assignee_id, and list_assignable_users with ignore_assigned_deals set to true lists other users by role.