MCP App Store
by KeenEthics
Zip icon

Zip MCP Server for Claude and ChatGPT

By Zip

Overview

Connect Claude to your Zip account to answer procurement questions in plain language, using your own data and your own Zip permissions. What you can do: - Check the status of purchase requests, who has approved them, and what is still pending - Look up vendors, agreements, purchase orders, invoices and credit memos - Review your pending approvals - Search the catalog and browse departments, locations, GL accounts, categories and other company settings - Inspect workflow configurations and intake forms to see how requests are routed

How to connect

  1. Claude: Open the Zip listing in the Claude connectors directory and choose Connect. Claude listing Zip exposes 70 tools here.
  2. ChatGPT: Open the Zip listing in the ChatGPT app directory and choose Connect. ChatGPT listing Zip exposes 69 tools here.

Tools

get_agreement

ChatGPT
Fetch a single agreement (contract) by GUID. Returns scalars: guid, name, agreement_number, currency, total_cents (the agreement amount in cents), start_date, execution_date, end_date, status, contract_status, display_renewal_status, renewal_status, type. Associations need expand; the less obvious ones are creator (the agreement owner), request_for_amendment (the parent request when type=AMENDMENT), requests (the requests this agreement was created from), and requests_for_master_contract (the parent requests when type=MSA). Note: the Zip API field 'original_request' is not directly returned — for MSA agreements use requests_for_master_contract[0], for AMENDMENT use request_for_amendment, otherwise use requests[0]. Renewal date fields are intentionally not returned (see FIO-1440): the underlying DB columns are misleadingly named and the CLM team is replacing them with properly-named fields behind the enable-new-renewal-fields flag. They will be added once that rename project is GA. Also not yet returned: auto_renews, opt_out_date, renewal_workflow, renewal_stakeholders, renewal_stakeholder_queues, cancellation_workflow_config (blocked on scheduled_renewal / queue zinterface registration), and custom attributes (FIAT). Associations are returned only when named in expand: attachments, categories, child_contracts, creator, department, purchase_orders, request_for_amendment, requests, requests_for_master_contract, subcategories, subsidiaries, vendor. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_attachment

ChatGPT
Get a single attachment (document) by GUID. Returns name, type, and status. MCP-callable subset of the GraphQL attachment read; permission-gated by can_user_view_attachment. Associations are returned only when named in expand: attachment_versions, contract, invoice, purchase_requisition, vendor. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_attachment_download_url

ChatGPT
Get a single-use download URL for the primary (latest) version of an attachment. Returns a time-limited, single-use URL (1 hour, one download — the URL expires after the first successful curl). Use the curl_command exactly as returned; the URL contains special characters that must be properly quoted. Get the attachment guid from any of these tools; attachments are only returned when expand names them: - get_agreement(guid, expand=["attachments"]) → attachments[].guid - get_invoice(guid, expand=["attachments"]) → attachments[].guid - get_purchase_requisition(guid, expand=["attachments"]) → attachments[].guid - list_agreements(..., expand=["attachments"]) → hits[].attachments[].guid - list_invoices(..., expand=["attachments"]) → hits[].attachments[].guid Example flow: 1. Call one of the above tools, with expand as shown, to get the attachment guids. 2. Call get_attachment_download_url(attachment_guid) to get a download URL. 3. curl -L the returned download_url to save the file locally. Args: attachment_guid: GUID of an Attachment (the primary/latest version is used) or of a specific AttachmentVersion (that version is used).

get_cash_back

ChatGPT
Get a single virtual card cash back by GUID. Returns the cash back's reference ID, amount and currency, transaction date, description, link, and any associated entity attributes (department, subsidiary, GL codes, location, custom fields).

get_category

ChatGPT
Fetch a single taxonomy node (category) by GUID. Returns basic info: name, description, status, type, prioritize_preferred_vendors.

get_change_order

ChatGPT
Get a single change order (purchase requisition of type CHANGE_ORDER) by GUID. Returns the change-order diff: changed header fields, plus created/deleted/updated PO line items, each with the attribute names and old/new values that changed. `field_type_options` selects which field groups to compute (ALL, PO_HEADER, LINE_ITEMS); defaults to ALL. Args: field_type_options: CoApiFieldTypeEnum names (ALL, PO_HEADER, LINE_ITEMS) selecting which change-order field groups to compute. Defaults to ALL.

get_context_library

ChatGPT
Fetch a single context library entry by GUID. Name latest_published_version in expand to get its purpose, instructions, and version number. Associations are returned only when named in expand: latest_published_version. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_credit_memo

ChatGPT
Get a single credit memo by GUID. Returns identity fields (guid, status, marked_as_used, created_at, updated_at). Does not include the underlying credit invoice, applied bill usages, or derived amounts; those require follow-up tooling.

get_currency

ChatGPT
Get a single currency by GUID. Returns its ISO code (name); name external_integration_datas in expand for its ERP system mappings. Returns the same ZCurrency shape as list_currencies. Associations are returned only when named in expand: external_integration_datas. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_department

ChatGPT
Get a single department by GUID with full details.

get_expense_category

ChatGPT
Fetch a single expense category by GUID. Expense categories tag spend with an ERP-aligned bucket (e.g. "Office Supplies", "Travel") and optionally carry a default GL account and subsidiary scope. Returns scalars (guid, name, source, description, active, hide_in_zip, created_at, updated_at). The linked GL account (gl_code), the subsidiaries the category is scoped to, and the ERP integration data rows that map it onto external_id values per integration (external_integration_datas) are available through expand. Associations are returned only when named in expand: external_integration_datas, gl_code, subsidiaries. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_field

ChatGPT
Get a single Zip configurable field (entity attribute config) by GUID. Returns the field's display name, internal name, entity type, status, and whether it's a Zip native field. Associations are returned only when named in expand: attribute_group_config, field_config_namespace. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_gl_account

ChatGPT
Get a single GL (general ledger) account by GUID. Associations are returned only when named in expand: external_integration_datas, subsidiaries. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_invoice

ChatGPT
Fetch a single BILL invoice by GUID. Returns core fields (id, invoice number, dates, amounts, currency, source, status flags). Audit-trail data and API-key-gated fields are not returned. Associations are returned only when named in expand: attachments, category, derived_matched_purchase_orders, invoice_department, line_items, location, payment_terms, posting_period, subcategory, subsidiary, vendor. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_item_account

ChatGPT
Get a single GL item (item_account) by GUID. Associations are returned only when named in expand: external_integration_datas, gl_code, subsidiaries. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_item_receipt

ChatGPT
Get a single item receipt (goods receipt) by GUID: status, receipt number, and received date. The matched purchase order, line items (with their attributes), and ERP external data are available through expand. Associations are returned only when named in expand: external_integration_datas, line_items, purchase_order. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_location

ChatGPT
Get a single location by GUID. Returns location name, status, and timestamps. Associations are returned only when named in expand: external_integration_datas. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_lookup

ChatGPT
Get a single lookup by GUID with full details: name, integration_name, integration_key, display_name_integration_fields, filters, settings_fields, visible.

get_lookup_option

ChatGPT
Get a single lookup option by GUID. Returns the option's name, key, and show_in_zip flag. Refuses GUIDs that aren't lookup options. Associations are returned only when named in expand: lookup. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_options_for_field

ChatGPT
List the available options for a Zip configurable field (entity attribute config). Returns each option's name, key, comment, score, color, link, external_id, and deletion/visibility flags. Optionally filter by a case-insensitive substring on the option name. Note: only options for fields in the caller's own organization are returned (no cross-org bypass, unlike the GraphQL OptionSearch endpoint). Args: entity_attribute_config_guid: GUID of the Zip configurable field (entity attribute config) whose options should be returned. search_text: Optional case-insensitive substring to filter option names by. first: Maximum number of options to return. Defaults to 20.

get_payment_terms

ChatGPT
Get a single payment terms entity by GUID. Returns name, net_period_in_days (Pay in), is_eom_terms (true for End of month; false or null for Net), and is_active. Associations are returned only when named in expand: external_integration_datas. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_payout

ChatGPT
Get a payout by GUID. Returns vendor, subsidiary, currency, fees, invoices, payment method, and key dates as ISO-8601 strings. Args: payout_guid: GUID of the payout or payment intent to fetch.

get_permission_group

ChatGPT
Get a single permission group ('tao_role' under the hood) by GUID. Returns the group's name, description, status, shared_condition_data, and permission-version metadata. Name resource_permissions in expand for the permissions wired to it (each with resource_type, level, and conditional condition_data). Vendor portal jurisdiction fields are included; the is_default and linked_scim_group fields shown in the Zip UI are not yet exposed here. Associations are returned only when named in expand: resource_permissions. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_product

ChatGPT
Fetch a single product by GUID. Includes the product's name, description, manufacturer info, UNSPSC code, type (GOODS, SERVICES, or SOFTWARE), and line type. Requires the configured organization to have the catalog API surface enabled. Associations are returned only when named in expand: commodity, gl_code, item_unit_of_measure. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_purchase_order

ChatGPT
Get a single purchase order by GUID. Returns scalar fields (po_number, status, source, currency, amount_cents, description, memo, payment_terms, version_number, order_date, sent_to_vendor, created_at, updated_at, external_id) and financial decimal-string fields (amount_str total, total_billed_amount_str, total_net_billed_amount_str, total_credited_amount_str, total_open_amount_str remaining balance, plus tax/shipping/misc). Custom attributes, change orders, and advance shipping notices are not yet exposed by this tool. Associations are returned only when named in expand: category, created_by_user, department, derived_matched_invoices, external_integration_datas, line_items, location, purchase_requisition, requester, subcategory, subsidiary, vendor. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_purchase_requisition

ChatGPT
Get a single purchase requisition by GUID with parity to the REST external API. Returns: core fields (name, description, status, priority, request number, timestamps, cycle/elapsed time, external id, request link, requested amount USD, payment method, is_existing_vendor); request attributes and total savings. Associations are returned only when named in expand: attachments, creator, department, engagement, external_integration_datas, gl_code, item_account, location, original_creator, owner, payment_terms_object, po_change_order, price_detail, purchase_category, purchase_order, purchase_subcategory, renewal_source_request, requester, subsidiary, vendor, vendor_portals, vendor_sourcing_datas, vendors, visible_purchase_order, workflow_config. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_purchase_requisition_audit_trail

ChatGPT
Get the audit trail for a purchase requisition: every recorded state transition (approval, rejection, cancellation, attribute change, etc.) with timestamp, actor, event type, description, and event-specific data. Includes available AI session metadata when the Superagent Observability flag is enabled for the organization. Requires the Audit Trail API feature for the organization; returns an empty list otherwise.

get_queue

ChatGPT
Get a single approval queue by GUID with full details: owners, members, default assignee, departments, parent queue (for subqueues), and subqueues (for parent queues). Use after list_queues to inspect a specific queue's membership and its subqueues. Associations are returned only when named in expand: assignee, conditions, department, employees, limit_conditions, owners, parent_job_title, sub_job_titles. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_subsidiary

ChatGPT
Get a single subsidiary by GUID with full details. Returns the same ZSubsidiary shape as list_subsidiaries: guid, name, display_name, country, currency, status.

get_top_up

ChatGPT
Get a single virtual card balance top up by GUID. Returns the top up's reference ID, amount and currency, transaction date, description, link, and any associated entity attributes.

get_transaction

ChatGPT
Fetch a single virtual card transaction by GUID. Returns core fields (id, reference id, dates, amounts, currency, merchant info, coding status, memo, links) and transaction-level coding attributes. The credit-card clearing GL code (credcard_gl_code) and per-line-item attributes (lines) are available through expand. The Brex/Stripe card source determines which fields are populated. Associations are returned only when named in expand: credcard_gl_code, lines. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_user

ChatGPT
Get a single user by GUID. Returns identity fields (email, first_name, last_name, employee_number, deactivated, created_at, updated_at). Department, subsidiary, and manager summaries, custom attributes, and external integration data are available through expand. Associations are returned only when named in expand: attributes, departments, external_integration_datas, managers, subsidiary. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_vendor

ChatGPT
Get a single vendor by GUID. Returns scalar fields (name, status, address, last_reviewed_at, created_at, updated_at, erp_currency). Many vendor attributes are accessible through `all_attributes: alternate_name, alternate_name_usage, currency, description, duns_number, legal_name, po_email, remittance_email, primary_phone_data, risk, tier, tax_id_type, tax_id_number, other_tax_id_type, other_tax_id_number, vat_number, website_link, categories, subsidiaries, department. ERP record IDs should be accessed through erp_id by passing the expected integration_type. When request_id is supplied, all_attributes returns the FM-merged request-scoped variant (mirrors the REST request_id query param). The deprecated external_id REST field is intentionally not exposed. Associations are returned only when named in expand: alternate_addresses, child_vendors, hq_address, parent_vendor, vendor_contacts, vendor_currencies, vendor_owners. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.` means x and everything beneath it, and `` means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_vendor_item

ChatGPT
Fetch a single vendor item by GUID. Vendor items are a vendor's offering of a product at a given price, optionally scoped to one or more purchasing groups (catalogs). The response includes status (DRAFT, ACTIVE, or DEACTIVATED), supplier part identifiers, lead time, and the preferred unit price and currency. The underlying product, the vendor, and the purchasing groups the item belongs to (catalogs) are available through expand. Requires the configured organization to have the catalog API surface enabled. Associations are returned only when named in expand: catalogs, product, vendor. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_vendor_item_group

ChatGPT
Fetch a single vendor item group by GUID. Vendor item groups bundle multiple vendor items together for purchasing, for example a kit of related parts or a PUNCHOUT-method bundle. The response includes the group's name, description, status (DRAFT, ACTIVE, or DEACTIVATED), purchasing method (CATALOG or PUNCHOUT), and supplier part identifiers. The catalogs (purchasing groups) it belongs to and its child memberships are available through expand; each child membership entry exposes the child vendor item's GUID, product name, and quantity. Requires the configured organization to have the catalog API surface enabled. Associations are returned only when named in expand: catalogs, child_memberships. Args: expand: Related objects to return, as dotted paths. Omit it to get only this object's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from the response rather than null.

get_workflow_config

ChatGPT
Fetch a single approval workflow config by GUID, its complete process-config graph, and the GUID of its editable (draft) version when workflow versioning is enabled. The graph includes parent groups and stages, child approvals and actions, parent-child membership, and outgoing dependency edges. For a child such as FINALIZE_DETAILS, use parent_process_config_guids to find the GROUP GUID accepted by workflow edge tools. Otherwise the returned config is editable directly. Named workflow and version statuses support polling asynchronous operations. The returned config also carries the workflow-level settings, so use this tool to read current values before changing them and to verify a change afterwards. That includes the notification, reminder, renewal, vendor-portal, sourcing, catalog, and agreement-visibility settings, plus the linked renewal workflow config, renewal cancellation workflow config, backup renewal queue, and filtered categories. It also carries receiving_tolerance with the over-receiving tolerance's enabled and percentage_tolerance, so writes to it are verifiable here. A null receiving_tolerance is ambiguous: it means either that the FLAG_PAYMENTS_OVER_RECEIVING_TOLERANCE flag is off for the organization or that no tolerance has ever been written. Attempting a write is the only way to tell those apart. On a BILL_APPROVAL child it also carries purchase_order_tolerance_configuration with its currency and all four per-type slots (po_tolerance, tax_tolerance, shipping_tolerance, misc_tolerance) hydrated, each with enabled, tolerance_metric, percentage_tolerance, and amount_cents_tolerance, so read current purchase-order tolerance values here before writing them. Null slots are unambiguous: a null tax_tolerance, shipping_tolerance, or misc_tolerance means no tolerance of that type is configured, since only po_tolerance is seeded when the PAYMENT_MASTER is created (PERCENTAGE, 0 percent, enabled). A null purchase_order_tolerance_configuration means something else: either this config is not the BILL_APPROVAL child, which is where the configuration hangs, or it is that child but predates the bootstrap. In the second case every purchase-order tolerance write against it fails until a configuration exists, so report the gap rather than retrying the write. On a GOODS_RECEIVING child it also carries process_rules: the one rule under that child's process group, whose guid is the process_rule_guid update_process_rule takes for the 3-way match policy, with the conditions currently attached to it. update_process_rule replaces the whole conditions list, so send every GUID read here plus any new one. An empty conditions list means no PO lines require 3-way match, not all of them. Other workflow types omit process_rules.

get_workflow_intake_form

ChatGPT
Fetch the requester primary intake form (questionnaire) for a workflow config: its sections, questions, question options, and attribute configs. Set include_conditions to also return each question's conditions (condition children nested one level). Mirrors the public Zip API GET /external_api/workflows/<guid>/intake_form endpoint. Args: include_conditions: When true, each question includes its conditions (with condition children nested one level). Defaults to false.

list_agreements

ChatGPT
List agreements (contracts) for the configured company. Optionally filter by vendor_guid, department_guid, subsidiary_guid, category_guid, subcategory_guid, creator_guid, owner_guid, agreement_number_str, renewal_status (list of symbolic names: UPCOMING, NOT_SCHEDULED, RENEWAL_IN_PROGRESS, CANCELLATION_IN_PROGRESS, RENEWAL_COMPLETE, CANCELLATION_COMPLETE, CANCELED), agreement_status (list of symbolic names: ACTIVE, INACTIVE, TERMINATED), total_cents_min/max (combine for a closed range), and end_date_after/before (ISO-8601 date; combine for a closed range). Sort by created_at (default), updated_at, end_date_or_max_date, or total_cents. Returns up to 50 per page. Each result has the same shape get_agreement returns. Renewal date fields and queue-related fields are not returned; see get_agreement's description for details. Associations on each row are returned only when named in expand: attachments, categories, child_contracts, creator, department, purchase_orders, request_for_amendment, requests, requests_for_master_contract, subcategories, subsidiaries, vendor. Args: vendor_guid: Filter to agreements with this vendor GUID. Exact match. department_guid: Filter to agreements in this department GUID. Exact match. Falls back to the vendor's primary department on agreements without a direct department. subsidiary_guid: Filter to agreements in this subsidiary GUID. Exact match. category_guid: Filter to agreements with this top-level category GUID. Exact match. subcategory_guid: Filter to agreements with this subcategory GUID. Exact match. creator_guid: Filter to agreements created by this user GUID. Exact match. owner_guid: Filter to agreements owned by this user GUID. Owner is currently resolved to the agreement's creator. agreement_number_str: Filter to a single agreement by its user-facing agreement number (as a string). renewal_status: Filter to agreements whose renewal status is in this set. Each entry is the symbolic name (e.g. "UPCOMING", "NOT_SCHEDULED", "RENEWAL_IN_PROGRESS", "CANCELLATION_IN_PROGRESS", "RENEWAL_COMPLETE", "CANCELLATION_COMPLETE", "CANCELED"). agreement_status: Filter to agreements whose status is in this set. Each entry is the symbolic name: "ACTIVE" (in effect today), "INACTIVE" (expired or not yet started), "TERMINATED". ACTIVE/INACTIVE derive from the start/end dates. TERMINATED is only available for companies with the terminated-status feature enabled; elsewhere a terminated agreement reads INACTIVE, since termination also moves the end date to the termination date. total_cents_min: Filter to agreements whose total amount in cents is greater than or equal to this value. Combine with `total_cents_max for a closed range. total_cents_max: Filter to agreements whose total amount in cents is less than or equal to this value. Combine with total_cents_min for a closed range. end_date_after: Filter to agreements whose end date is on or after this date. ISO-8601 date string (e.g. "2026-01-15"). Combine with end_date_before for a closed range. Use a date without a time component. end_date_before: Filter to agreements whose end date is on or before this date. ISO-8601 date string. Combine with end_date_after for a closed range. sort_by: Field to sort by. One of: created_at (default), updated_at, end_date_or_max_date, total_cents. sort_order: Sort direction. One of: asc, desc (default). expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.` means x and everything beneath it, and ` means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size. Changed 2026-11-11: Description changed for tool 'list_agreements'.

list_approvals

ChatGPT
List approval nodes for the configured company. Optionally filter by request_number, node_status_id, needs_my_actions_only (true = only the caller's own approvals awaiting action, i.e. in progress or paused, same rule and request types as the Home 'Needs my action' tab), request_types (symbolic PurchaseRequisitionType names, e.g. PURCHASE_REQUEST, BILL, SOURCING_PROJECT), config_types (symbolic ProcessConfigType names), request_created_after/before (ISO-8601 date; combine for a closed range), and due_after/before (ISO-8601 date; combine for a closed range). Sort by created_at (default), updated_at, due_date, request_initiated_at, or request_number. Returns up to 50 per page; follow the cursor until it is empty before counting or summarizing. Each hit is a ZProcessNode with timing fields (start/complete/due as epoch seconds, duration and pause seconds), display_status, the assignee, the originating request, and the subtask checklist (tasks). Results are already scoped to approvals the caller can see (approver, queue member, or org-wide approval viewer). Associations on each row are returned only when named in expand: assignee, request_for_actionable_process_node, tasks. Args: request_number: Filter to the approval whose request has this user-facing request number. Exact match. node_status_id: Filter to approvals with this custom node-status GUID (or built-in status key). Exact match. needs_my_actions_only: When true, only the caller's own approvals that are awaiting action: nodes in progress or paused where the caller is the approver or an active-subtask assignee. Same rule as the web Home "Needs my action" tab, so it covers the request types that tab shows (including invoice exception tasks) unless request_types is given. Excludes upcoming, approved, rejected and other finished approvals. request_types: Filter to approvals whose request is one of these types. Each entry is the symbolic name, e.g. "PURCHASE_REQUEST", "BILL", "SOURCING_PROJECT", "PAYOUT_GROUP", "CHANGE_ORDER". config_types: Filter to approvals whose workflow config is one of these types. Each entry is the symbolic ProcessConfigType name. request_created_after: Only approvals whose request was created on or after this date. ISO-8601 date (e.g. "2026-01-15"). Combine with request_created_before for a closed range. request_created_before: Only approvals whose request was created on or before this date. ISO-8601 date. Combine with request_created_after for a closed range. due_after: Only approvals due on or after this date. ISO-8601 date. Combine with due_before for a closed range. due_before: Only approvals due on or before this date. ISO-8601 date. Combine with due_after for a closed range. sort_by: Field to sort by. One of: created_at (default), updated_at, due_date, request_initiated_at, request_number. sort_order: Sort direction. One of: asc, desc (default). expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_change_orders

ChatGPT
List submitted change orders (purchase requisitions of type CHANGE_ORDER) for the configured company. Returns up to 50 hits per page, ordered newest-first; pass the returned cursor to page further. Each hit is the same change-order diff shape `get_change_order returns (changed header fields and created/deleted/updated PO line items). Args: field_type_options: CoApiFieldTypeEnum names (ALL, PO_HEADER, LINE_ITEMS) selecting which change-order field groups to compute. Defaults to ALL. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_context_libraries

ChatGPT
List context library entries for the configured company. Optionally filter by status. Returns up to 50 per page. Hits are hydrated to the same ZAiCompanyContext payload that get_context_library returns. Associations on each row are returned only when named in expand: latest_published_version. Args: status: Filter by status. One of "ACTIVE" or "ARCHIVED". Omit to return both. expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_credit_memos

ChatGPT
List credit memos for the configured company. Optionally filter by vendor name (case-insensitive whole-token / token-prefix match — not arbitrary substrings), exact invoice number, or availability. Results are limited to credit memos you have permission to view based on subsidiary, department, and assignee. Each result has the same shape get_credit_memo returns. Returns up to 50 per page, sorted by created_at descending. Args: vendor_query: Optional case-insensitive token / phrase match on the associated vendor's name. Whole tokens (or token prefixes) only — arbitrary substrings won't match (e.g. `"cme" will NOT match "Acme Inc"). Omit to return credit memos for all vendors. invoice_number: Optional exact-match filter on the underlying credit invoice's invoice_number. Omit to return all matching credit memos. is_credit_available: Optional boolean filter. Set True to return only credit memos with an available balance, False for fully-used credit memos. Omit to return both. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size. Changed 2026-11-11: Description changed for tool 'list_credit_memos'.

list_currencies

ChatGPT
List the supported currencies for the configured company. Each currency includes its ISO code (as the name field) and any external integration data (e.g. ERP system mappings) under external_integration_datas. Optionally filter by a case-insensitive substring on the ISO code; total reflects the post-filter match count, not the unfiltered company total. Loosely mirrors GET /external_api/currencies, with two deliberate divergences: the per-item external_data list is renamed to external_integration_datas and exposes the full ZExternalIntegrationData shape (guid, erp_object_type, external_id, external_link, source) rather than the flat {id, source} pair the REST endpoint returns; and the redundant size key is dropped (it always equaled total). Args: iso_code: Optional case-insensitive substring filter on the ISO code (e.g. 'USD', 'EUR'). Omit to return every supported currency. When set, the response 'total' is the post-filter match count, not the unfiltered company total.

list_departments

ChatGPT
List active departments for the configured company. Optionally filter by name (case-insensitive substring), hide_in_zip (visibility), or external_id (ERP-prefixed identifier exact match). Sort by created_at or updated_at. Returns up to 50 per page. Args: name: Case-insensitive substring filter on the department's display name. Omit to match all names. hide_in_zip: When True, return only departments hidden from the Zip UI; when False, return only visible departments. Omit to return both. external_id: Filter to departments whose `external_ids array contains this value. The value is the ERP-prefixed identifier stored on the entity (e.g. netsuite:abc123); the prefix is required for an exact match. sort_by: Field to sort by. One of: created_at (default), updated_at. sort_order: Sort direction. One of: asc, desc (default). cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_expense_categories

ChatGPT
List expense categories for the configured company. Optionally filter by name (case-insensitive substring match), active (True / False), and external_ids (matches categories whose ERP external_id is in the supplied list). Returns up to 50 per page. Each hit has the same shape get_expense_category returns. Associations on each row are returned only when named in expand: external_integration_datas, gl_code, subsidiaries. Args: name: Case-insensitive substring filter on the expense category name. Omit to return all matching categories. Note: the underlying ES field is KEYWORD-mapped, so the substring uses wildcard matching. active: If True, return only active categories. If False, return only inactive categories. Omit to return both. external_ids: Filter to categories whose ERP external_id is in this list. Exact-match on the keyword index. Omit to return categories regardless of their external_id. expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size. Changed 2026-11-11: Description changed for tool 'list_expense_categories'.

list_fields

ChatGPT
List Zip configurable fields (entity attribute configs) for the configured company. Each hit returns the field's GUID, namespace-prefixed display name, and namespace integer code. Optionally filter by namespace (REQUEST, VENDOR, CONTRACT, etc.) and/or a case-insensitive substring on the display name. Note: results are scoped to the caller's organization. Returns up to 50 per page. Args: namespace: Filter to fields belonging to a single ZipFieldConfigNamespace. Pass the enum name as a string. Valid values: REQUEST, VENDOR, CONTRACT, INVOICE, PAYMENT, REQUEST_LINE, VENDOR_CREDIT, VENDOR_CREDIT_LINE (see ZipFieldConfigNamespace). Omit to return fields across all namespaces. name: Case-insensitive substring filter on the field's display name (matched against the namespace-prefixed display name, e.g. "(Request) Vendor category"). Omit to return all fields. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_gl_accounts

ChatGPT
List GL (general ledger) accounts for the configured company. Optionally filter by name (case-insensitive substring), account_types, include_inactive, and include_hidden. Hits have the same shape get_gl_account returns. Returns up to 50 per page. Associations on each row are returned only when named in expand: external_integration_datas, subsidiaries. Args: name: Case-insensitive substring filter on the GL account display name (or canonical name when no display_name is set). Omit to return all matching accounts. account_types: Filter to GL accounts whose account_type is in this list (e.g. ["EXPENSE", "LIABILITY"]). Omit to return all account types. include_inactive: If False (default), return only active accounts. Set True to include inactive accounts as well. include_hidden: If True (default), return GL accounts regardless of their hide_in_zip flag. Set False to exclude accounts hidden from the Zip UI. expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_invoices

ChatGPT
List BILL invoices for the configured company. BILL is the only invoice type returned (CREDIT and PROFORMA are excluded). Filter by vendor, subsidiary, line-item department or GL code, exact invoice number, BillStatus name, invoiced-amount range (in cents, not dollars), approval timestamp range, and update timestamp range. Returns up to 50 per page. Each result has the same shape get_invoice returns. Results are limited to invoices you have permission to view. total counts matches before permission filtering, so it can exceed the number of returned results; treat it as an upper bound. Associations on each row are returned only when named in expand: attachments, category, derived_matched_purchase_orders, invoice_department, line_items, location, payment_terms, posting_period, subcategory, subsidiary, vendor. Args: vendor_id: Filter to invoices for this vendor GUID. Exact match. status: Filter to invoices whose derived bill status is in this set. Each entry must be a `BillStatus name (e.g. "INITIAL", "PENDING_APPROVAL", "APPROVED", "PAID"). The full set is enumerated by BillStatus in thriftgen/models/bill/ttypes.py and reflected in the validator error when an invalid name is passed. subsidiary_id: Filter to invoices on this subsidiary GUID. Exact match. department_id: Filter to invoices that have at least one line item on this department GUID. Exact match. gl_code_id: Filter to invoices that have at least one line item on this GL code GUID. Exact match. invoice_number: Filter to a single invoice by its vendor-supplied invoice number. Exact match (keyword). invoice_amount_min: Filter to invoices whose invoiced amount in cents is >= this value. invoice_amount_max: Filter to invoices whose invoiced amount in cents is <= this value. approved_after: Filter to invoices approved (bill_approved_datetime) at or after this ISO-8601 timestamp. Combine with approved_before for a closed range. approved_before: Filter to invoices approved at or before this ISO-8601 timestamp. Combine with approved_after for a closed range. updated_after: Filter to invoices last updated at or after this ISO-8601 timestamp. Combine with updated_before for a closed range. updated_before: Filter to invoices last updated at or before this ISO-8601 timestamp. Combine with updated_after for a closed range. expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.` means x and everything beneath it, and ` means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size. Changed 2026-11-11: Description changed for tool 'list_invoices'. Changed 2026-11-09: Param 'status' on tool 'list_invoices' changed shape.

list_item_accounts

ChatGPT
List GL items (item_accounts) for the configured company. Optionally filter by name (case-insensitive substring) and/or active status. Returns up to 50 per page. Args: name: Case-insensitive substring filter on the GL item's name. Omit to skip the name filter. active: If true, only return active GL items. If false, only return inactive ones. Omit to return both. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_item_receipts

ChatGPT
List item receipts (goods receipts) for the configured company. Optionally filter by status (DRAFT, SUBMITTED, VOID, ARCHIVED) and created/updated date ranges; sort by created_at or updated_at. Each result has the same shape get_item_receipt returns. Returns up to 100 per page, limited to receipts you have permission to view. total counts matches before permission filtering, so it can exceed the number of returned results; treat it as an upper bound. Associations on each row are returned only when named in expand: external_integration_datas, line_items, purchase_order. Args: status: Filter to goods receipts whose status is in this set. Each entry must be a GoodsReceiptStatus name (e.g. "DRAFT", "SUBMITTED", "VOID", "ARCHIVED"). Omit to return all statuses. created_after: Return receipts created at or after this ISO-8601 timestamp. Combine with created_before for a closed range. created_before: Return receipts created at or before this ISO-8601 timestamp. Combine with created_after for a closed range. updated_after: Return receipts last updated at or after this ISO-8601 timestamp. Combine with updated_before for a closed range. updated_before: Return receipts last updated at or before this ISO-8601 timestamp. Combine with updated_after for a closed range. sort_by: Field to sort by. One of: created_at (default), updated_at. sort_order: Sort direction. One of: asc, desc (default). expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size. Changed 2026-11-11: Description changed for tool 'list_item_receipts'.

list_locations

ChatGPT
List locations for the configured company. Optionally filter by name (case-insensitive substring) or status (INITIAL, ACTIVE, DELETED). Sort by created_at or updated_at. Returns up to 50 per page. Associations on each row are returned only when named in expand: external_integration_datas. Args: name: Case-insensitive substring filter on the location's display name. Omit to match all names. status: Filter to locations with this LocationStatus. One of: INITIAL, ACTIVE, DELETED. Omit to match all statuses. sort_by: Field to sort by. One of: created_at (default), updated_at. sort_order: Sort direction. One of: asc, desc (default). expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_lookup_options

ChatGPT
List lookup options for the configured company. Deleted options are excluded. Optionally filter by case-insensitive name substring, exact integration key, parent lookup GUID, or show_in_zip. Each hit is a full lookup-option payload matching the get_lookup_option shape. Returns up to 50 per page. Associations on each row are returned only when named in expand: lookup. Args: name: Case-insensitive substring filter on the option's display name. key: Exact-match filter on the option's integration key. lookup_id: GUID of the parent Lookup. Omit to return options across every lookup. show_in_zip: Filter to options visible in the Zip UI (true) or hidden (false). Omit for both. expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_lookups

ChatGPT
List lookups for the configured company. Optionally filter by name substring (case-insensitive), integration_name, integration_key, or show_in_zip. Returns up to 50 per page. Args: name: Case-insensitive substring filter on the lookup's name. Omit to skip the name filter. integration_name: Exact-match filter on the integration name (e.g. "NETSUITE", "WORKDAY"). Case-insensitive — matched against the lowercased index. Omit for all integrations. integration_key: Exact-match filter on the integration key. Useful for finding a specific lookup tied to an ERP object. show_in_zip: Filter to lookups visible in the Zip UI (true) or hidden (false). Omit for both. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_payment_terms

ChatGPT
List payment terms for the configured company. Optionally filter by name (case-insensitive token match) or is_active. Returns hydrated payment terms with name, net_period_in_days (Pay in), is_eom_terms (true for End of month; false or null for Net), and is_active. Returns up to 50 per page. Associations on each row are returned only when named in expand: external_integration_datas. Args: name: Case-insensitive token match on the payment terms name (e.g. "Net 30"). Omit to match all names. is_active: When True, return only active payment terms; when False, return only inactive. Omit to return both. expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_payouts

ChatGPT
List payouts for the configured company. Optional filters: paid_on_date, payment_provider_creation_time, reversed_on_date (all ISO-8601 datetime strings); payment_method_type, one of: ZIP_CLEARING_ACCOUNT (paid from the company's Zip clearing balance), ZIP_VIRTUAL_CARD (paid via a Zip-issued virtual card), USER_OFFLINE (recorded as paid outside Zip); subsidiary_id. Page size capped at 50. Args: paid_on_date_after: Filter to payouts paid on or after this ISO-8601 datetime. paid_on_date_before: Filter to payouts paid on or before this ISO-8601 datetime. payment_provider_creation_time_after: Filter to payouts whose payment provider creation time is on or after this ISO-8601 datetime. payment_provider_creation_time_before: Filter to payouts whose payment provider creation time is on or before this ISO-8601 datetime. reversed_on_date_after: Filter to payouts reversed on or after this ISO-8601 datetime. reversed_on_date_before: Filter to payouts reversed on or before this ISO-8601 datetime. payment_method_type: Filter to payouts of this payment method type. One of: ZIP_CLEARING_ACCOUNT (paid from the company's Zip clearing balance), ZIP_VIRTUAL_CARD (paid via a Zip-issued virtual card), USER_OFFLINE (recorded as paid outside Zip). subsidiary_id: Filter to payouts for the subsidiary with this GUID. page_token: Opaque cursor returned by a previous call as `next_page_token`. page_size: Maximum number of payouts to return. Capped at 50.

list_permission_groups

ChatGPT
List visible permission groups ('tao_role' under the hood) for the configured company. Returns the same set the Zip 'Manage users and permission groups' page shows: status VISIBLE, owned by the caller's company. Vendor portal groups are not included. Each group includes guid, name, description, status, shared_condition_data, and permission-version metadata. The is_default and linked_scim_group fields shown in the Zip UI are not yet exposed here -- use the UI for those.

list_primary_categories

ChatGPT
List the top-level (primary) taxonomy nodes for a company. Excludes INACTIVE and soft-deleted nodes; categories with no status set are treated as active and are returned. Optionally filter by type and by name (case-insensitive substring match). Valid types: PURCHASE_CATEGORY, REQUEST_CATEGORY, QUESTION_CATEGORY, CANONICAL_CATEGORY, CUSTOM_CATEGORY.

list_purchase_orders

ChatGPT
List purchase orders for the configured company. Filter by vendor GUID, last-update timestamp range, and open-amount range (USD, decimal-string dollars). Returns up to 50 hits per page. Each hit has the same `ZPurchaseOrder shape get_purchase_order returns. Associations on each row are returned only when named in expand: category, created_by_user, department, derived_matched_invoices, external_integration_datas, line_items, location, purchase_requisition, requester, subcategory, subsidiary, vendor. Args: vendor_guid: Filter to purchase orders for this vendor GUID. Exact match. created_after: Filter to POs created at or after this ISO-8601 timestamp (e.g. "2026-01-15T00:00:00Z"). Combine with created_before for a closed range. created_before: Filter to POs created at or before this ISO-8601 timestamp. Combine with created_after for a closed range. last_updated_after: Filter to POs whose attributes were last updated at or after this ISO-8601 timestamp (e.g. "2026-01-15T00:00:00Z"). Combine with last_updated_before for a closed range. last_updated_before: Filter to POs whose attributes were last updated at or before this ISO-8601 timestamp. Combine with last_updated_after for a closed range. open_amount_greater_than: Filter to POs whose open amount (USD) is greater than this decimal-string dollar amount (e.g. "100.50"). Combine with open_amount_less_than for a closed range. open_amount_less_than: Filter to POs whose open amount (USD) is less than this decimal-string dollar amount. Combine with open_amount_greater_than for a closed range. sort_by: Field to sort by. One of: created_at (default), updated_at. When filtering by last_updated_after / last_updated_before, pass updated_at so the sort field matches the filter field. sort_order: Sort direction. One of: asc, desc (default). expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.` means x and everything beneath it, and ` means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_purchase_requisitions

ChatGPT
List purchase requisitions for the configured company. Filter by approval status (pass approval_status=["AWAITING_APPROVAL"] for pending requests), request type, requester, vendor, department, subsidiary, priority, request number, and creation/update timestamp ranges. Sort by creation or update time. Args: approval_status: Filter to requisitions whose approval status is in this set. Each entry is an ApprovalStatus name: DRAFT, AWAITING_APPROVAL, REJECTED, APPROVED, CANCELED, CLOSED, PAUSED. The Zip UI labels AWAITING_APPROVAL as "Requested", so a request for pending / open / in-flight / awaiting-approval requisitions maps to ["AWAITING_APPROVAL"]. request_type: Filter to requisitions of these types. Each entry is a PurchaseRequisitionType name, e.g. "PURCHASE_REQUEST", "BILL", "SECURITY_REVIEW", "VENDOR_ONBOARDING", "SOURCING_REQUEST", "PO_CHANGE_ORDER", "CHANGE_ORDER". Requisitions with no type recorded do not match any request_type filter value. requester_guid: Filter to requisitions whose requester (creator) has this user GUID. Exact match. requester_name_key: Filter by requester full name (case-insensitive substring match, so "Alice" matches "Alice Smith"). Prefer `requester_guid when known. vendor_guid: Filter to requisitions with this vendor GUID. Exact match against the requisition's displayed (primary) vendor only — requisitions where this vendor is a secondary or sourcing vendor are not matched. vendor_name_key: Filter by vendor name (case-insensitive substring match, so "Arista" matches "Arista Networks Solutions, Inc"). Prefer vendor_guid when known. department_guid: Filter to requisitions in this department GUID. Exact match. department_name_key: Filter by department name (case-insensitive substring match). subsidiary_guid: Filter to requisitions in this subsidiary GUID. Exact match. priority: Filter to requisitions whose priority is in this set (integer enum values). request_number_str: Filter to a single requisition by its user-facing request number (as a string). created_after: Filter to requisitions created at or after this timestamp. ISO-8601 string (e.g. "2026-01-15T00:00:00Z"). Combine with created_before for a closed range. created_before: Filter to requisitions created at or before this timestamp. ISO-8601 string. Combine with created_after for a closed range. last_updated_after: Filter to requisitions whose attributes were last updated at or after this timestamp. ISO-8601 string. Combine with last_updated_before for a closed range. last_updated_before: Filter to requisitions whose attributes were last updated at or before this timestamp. ISO-8601 string. Combine with last_updated_after for a closed range. sort_by: Field to sort by. One of: created_at (default), updated_at, initiated_or_created_at_date, last_attribute_updated_at, request_number_str. sort_order: Sort direction. One of: asc, desc (default). cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size. Changed 2026-11-11: Description changed for tool 'list_purchase_requisitions'.

list_questionnaire_responses_for_requisition

ChatGPT
List questionnaire responses attached to a purchase requisition. Returns the questionnaire name, status, and completion timestamp for each response. The detailed question-response tree is not included in this tool's output. Associations on each row are returned only when named in expand: vendor. Args: requisition_guid: The purchase requisition GUID to filter by. Returns all questionnaire responses associated with this requisition. expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.* means x and everything beneath it, and * means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null.

list_queues

ChatGPT
List active approval queues for the configured company. Optionally filter by name substring, by department GUID, or by top-level/subqueue status. Always excludes inactive queues and non-queue employee_job_title rows. Returns summary fields per row; call get_queue for full details. Returns up to 50 per page. Args: name: Case-insensitive substring filter on the queue's name. Omit to skip the name filter. department_guid: Return only queues attached to this department. Omit for all departments. top_level_only: If True, return only top-level queues (no parent). If False, return only subqueues. Omit for both. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_request_workflows

ChatGPT
List the company's purchase request approval workflows. Returns every PURCHASE_REQUEST workflow config regardless of status (including archived) plus its name and status. Mirrors the public Zip API GET /external_api/workflows endpoint. For the broader workflows-settings view (all workflow types, live statuses only, templates excluded) use list_workflow_configs instead.

list_subcategories

ChatGPT
List direct subcategories (one level down, not the full subtree) of a taxonomy node. Excludes INACTIVE and soft-deleted nodes; categories with no status set are treated as active and are returned. Optionally filter by type. Valid types: PURCHASE_CATEGORY, REQUEST_CATEGORY, QUESTION_CATEGORY, CANONICAL_CATEGORY, CUSTOM_CATEGORY. Use list_primary_categories for top-level categories. Call this once per level to walk a multi-level category tree.

list_subsidiaries

ChatGPT
List subsidiaries for the configured company. Optionally filter by name (case-insensitive substring match on the display name or canonical name) and include_inactive (default False — only active subsidiaries are returned, matching the REST search default). Returns up to 50 per page. Each hit is a full ZSubsidiary with the same scalar fields get_subsidiary returns: guid, name, display_name, country, currency, status. Args: name: Filter to subsidiaries whose display name (or canonical name when no display_name is set) contains this string. Case-insensitive substring match. Omit to list all subsidiaries. include_inactive: When `False (the default), only active subsidiaries are returned — matches the REST SubsidiaryApiSearchHandler default. When True, both active and inactive subsidiaries are returned. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_users

ChatGPT
List users in the configured company. Optionally filter by exact email. Excludes Zip employees and contractors. Returns up to 50 per page, sorted by created_at descending. Args: email: Optional exact-match filter on the user's email address. Omit to return users regardless of email. deactivated: When omitted, returns active users only (matches the REST handler default). Set `True to return only deactivated users; set False to return only active users. sort_by: Field to sort by. One of: created_at (default), updated_at, email, employee_number. sort_order: Sort direction. One of: asc, desc (default). cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size.

list_vendors

ChatGPT
List vendors for the configured company. Optionally filter by name (trimmed, case-insensitive substring match), or primary, legal, and alternate name tokens/prefixes when include_name_variants is true; category_guid, subcategory_guid, custom_category_guid (including descendants), department_guid, subsidiary_guid, status (list of VendorStatus enum names: ACTIVE, INITIAL, DRAFT, DELETED; raises an error if status filtering is unavailable for the company), and a last_updated_after / last_updated_before timestamp range. Sort by created_at, updated_at, or last_updated. Returns the same shape as `get_vendor for each hit. Returns up to 50 per page. Associations on each row are returned only when named in expand: alternate_addresses, child_vendors, hq_address, parent_vendor, vendor_contacts, vendor_currencies, vendor_owners. Args: name: Filter by vendor name. Trimmed and matched as a case-insensitive substring, so "Arista" matches "Arista Networks Solutions, Inc". include_name_variants: Match name tokens and prefixes across the primary, legal, and alternate names instead of the primary-name substring match. category_guid: GUID of a legacy vendor category. subcategory_guid: GUID of a legacy vendor subcategory. custom_category_guid: GUID of a custom taxonomy category at any depth. Includes vendors assigned to descendant categories. department_guid: Filter to vendors whose primary department has this GUID. Exact match. subsidiary_guid: Filter to vendors associated with this subsidiary GUID. status: Filter to vendors whose status is in this list. Each entry must be a VendorStatus enum name (e.g. ACTIVE, INITIAL, DRAFT, DELETED). Raises an error if status filtering is unavailable for the company. An empty list applies no filter. last_updated_after: Filter to vendors whose last_updated timestamp is at or after this value. ISO-8601 string (e.g. "2026-01-15T00:00:00Z"). Combine with last_updated_before for a closed range. last_updated_before: Filter to vendors whose last_updated timestamp is at or before this value. ISO-8601 string. Combine with last_updated_after for a closed range. sort_by: Field to sort by. One of: created_at, updated_at, last_updated (default). sort_order: Sort direction. One of: asc, desc (default). expand: Related objects to return on each row, as dotted paths. Omit it to get only each row's own fields, with no associations. x.` means x and everything beneath it, and ` means the whole tree. Selecting a nested path implies its parents. An association you do not select is absent from each row rather than null. cursor: Opaque pagination cursor. Omit for the first page; to fetch the next page, pass back the cursor` value from the previous response. page_size: Maximum rows to return for this page (1-100). Defaults to the endpoint's configured page size. Changed 2026-11-11: Description changed for tool 'list_vendors'.

list_workflow_configs

ChatGPT
List approval workflow configs for the configured company. Mirrors the workflows page in company settings. Supports pagination and filtering by workflow config type, e.g. PURCHASE_REQUEST, CHANGE_ORDER, or case-insensitive substring match.

search_catalog

ChatGPT
Search the catalog for vendor items and vendor item groups by free-text query and / or filters. Returns two typed result lists: vendor_items and vendor_item_groups. Supports filtering by category, subcategory, vendor, subsidiary, status (default ACTIVE), and vendor item group purchasing method (default CATALOG). Pagination via page (1-indexed) and page_size. The metadata field carries total page and per-type counts. Requires the configured organization to have the catalog API surface enabled. Args: search_text: Free-text search query. Empty / omitted returns all items matching the other filters. page: 1-indexed page number. Defaults to 1. page_size: Page size. Defaults to 20. sort: Field to sort by. Default is search relevance. order: Sort order. One of: asc, desc. Only used when sort is provided. aggregation_only: When true, returns only metadata (counts) and no result rows. Defaults to false. category_guids: Restrict results to vendor items / groups in any of these categories. subcategory_guids: Restrict results to vendor items / groups in any of these subcategories. Merged with category_guids on the backend. vendor_guids: Restrict results to vendor items / groups owned by any of these vendors. subsidiary_guids: Restrict results to subsidiary GUIDs. statuses: Vendor item / group statuses to include. Defaults to ['ACTIVE']. One or more of DRAFT, ACTIVE, DEACTIVATED. vendor_item_group_purchasing_methods: Restrict vendor item groups to these purchasing methods. Defaults to ['CATALOG']. One or more of CATALOG, PUNCHOUT. object_types: Result entity types. Defaults to both. One or more of vendor_item, vendor_item_group.

whoami

ChatGPT
Return the authenticated user's GUID, email, organization GUID, and company name.

get_agreement

Claude

get_attachment

Claude

get_attachment_download_url

Claude

get_cash_back

Claude

get_category

Claude

get_change_order

Claude

get_context_library

Claude

get_credit_memo

Claude

get_currency

Claude

get_department

Claude

get_expense_category

Claude

get_field

Claude

get_gl_account

Claude

get_invoice

Claude

get_item_account

Claude

get_item_receipt

Claude

get_location

Claude

get_lookup

Claude

get_lookup_option

Claude

get_options_for_field

Claude

get_payment_terms

Claude

get_payout

Claude

get_permission_group

Claude

get_product

Claude

get_purchase_order

Claude

get_purchase_requisition

Claude

get_purchase_requisition_audit_trail

Claude

get_queue

Claude

get_subsidiary

Claude

get_top_up

Claude

get_transaction

Claude

get_user

Claude

get_vendor

Claude

get_vendor_item

Claude

get_vendor_item_group

Claude

get_workflow_config

Claude

get_workflow_intake_form

Claude

join_company

Claude

list_agreements

Claude

list_approvals

Claude

list_change_orders

Claude

list_context_libraries

Claude

list_credit_memos

Claude

list_currencies

Claude

list_departments

Claude

list_expense_categories

Claude

list_fields

Claude

list_gl_accounts

Claude

list_invoices

Claude

list_item_accounts

Claude

list_item_receipts

Claude

list_locations

Claude

list_lookup_options

Claude

list_lookups

Claude

list_payment_terms

Claude

list_payouts

Claude

list_permission_groups

Claude

list_primary_categories

Claude

list_purchase_orders

Claude

list_purchase_requisitions

Claude

list_questionnaire_responses_for_requisition

Claude

list_queues

Claude

list_request_workflows

Claude

list_subcategories

Claude

list_subsidiaries

Claude

list_users

Claude

list_vendors

Claude

list_workflow_configs

Claude

search_catalog

Claude

whoami

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/zip"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/zip)