Tool: Agiloft — Zelaxy Docs

Full Agiloft CLM integration for managing contracts and records — create, read, update, delete, and search across any table in your knowledge base. Ideal for automating contract lifecycle operations, syncing CLM data with other systems, and building AI-powered contract workflows.

Overview

Property	Value
Type	`agiloft`
Category	Tool — CRM / Contract Lifecycle Management
Auth	Username + Password (session token exchange via Agiloft REST login endpoint)

Operations

Operation	Tool ID	Description
Create Record	`agiloft_create_record`	Create a new record in an Agiloft table
Read Record	`agiloft_read_record`	Fetch a single record by ID from an Agiloft table
Update Record	`agiloft_update_record`	Modify field values on an existing record
Delete Record	`agiloft_delete_record`	Permanently delete a record by ID
Query Records	`agiloft_query_records`	Search for records using Agiloft query syntax with optional pagination

Configuration

Setting	Type	Required	Description
`instanceUrl`	`string`	Yes	Base URL of your Agiloft instance, e.g. `https://mycompany.agiloft.com`. No trailing slash.
`knowledgeBase`	`string`	Yes	Name of the Agiloft knowledge base to operate against.
`login`	`string`	Yes	Agiloft username used to authenticate. Store this as a secret and reference it with `{{AGILOFT_LOGIN}}`.
`password`	`string`	Yes	Agiloft password. Store this as a secret and reference it with `{{AGILOFT_PASSWORD}}`.
`table`	`string`	Yes	Table (object type) to operate on, e.g. `Contract`, `contacts.employees`. Supports dotted sub-table paths.
`recordId`	`string`	Yes (Read / Update / Delete)	ID of the record to read, update, or delete. Not used for Create or Query operations.
`data`	`string`	Yes (Create / Update)	JSON object of field names and values to write, e.g. `{"first_name": "Jane", "status": "Active"}`. Must be valid JSON.
`query`	`string`	Yes (Query Records)	Agiloft query expression to filter results, e.g. `status='Active'`.
`fields`	`string`	No (Read / Query)	Comma-separated list of field names to include in the response. Omit to return all fields.
`page`	`string`	No (Query Records)	Zero-based page number for paginated results. Defaults to `0`.
`limit`	`string`	No (Query Records)	Maximum number of records to return per page. Defaults to `25`.

Outputs

Create Record / Read Record / Update Record

Field	Type	Description
`id`	`string`	ID of the created, read, or updated record.
`fields`	`json`	Full field values of the record as returned by Agiloft.

Delete Record

Field	Type	Description
`id`	`string`	ID of the deleted record (echoed from the input).
`deleted`	`boolean`	`true` when the deletion succeeded; `false` on failure.

Query Records

Field	Type	Description
`records`	`json`	Array of matching record objects, each containing field names and values.
`totalCount`	`number`	Total number of records matching the query (may exceed the current page).
`page`	`number`	Current page number (zero-based).
`limit`	`number`	Number of records returned per page.

Example

[Starter] → [Agiloft: Query Records] → [Agent: Summarize contracts] → [Agiloft: Update Record]

Connect a scheduled Starter to an Agiloft Query Records block configured with table = Contract and query = status='Under Review' using credentials {{AGILOFT_LOGIN}} and {{AGILOFT_PASSWORD}}. Pass {{agiloft.records}} to an Agent that summarizes each contract and determines the next action. Feed the Agent's decision back into an Update Record block, setting recordId = {{starter.contractId}} and data = {"status": "Approved", "notes": "{{agent.content}}"} to write the result back to Agiloft.

Tips

Credentials are exchanged for a session token on each call. The integration logs in, performs the operation, and logs out automatically. Store login and password as workspace secrets and reference them as {{AGILOFT_LOGIN}} and {{AGILOFT_PASSWORD}} to keep them out of your workflow definition.
Table names are case-sensitive and must match the object type name in your Agiloft knowledge base exactly. Sub-tables use dotted notation, e.g. contacts.employees.
data must be valid JSON. For Create and Update operations the block will return an error (Invalid JSON in data parameter) if the string cannot be parsed. Use a JSON object literal or reference structured output from an upstream block with {{agent.content}}.
Use fields to limit response size. When reading or querying large records, pass a comma-separated list to fields (e.g. id,status,contract_value) instead of fetching the full object, which can be large.
Paginate large query result sets. The default page size is 25. For bulk operations, loop over pages by incrementing page starting from 0 until records returns fewer items than limit.
Query syntax follows Agiloft's EWSEARCH expression language. Simple equality looks like status='Active'; use AND/OR to combine conditions, e.g. status='Active' AND contract_type='MSA'.

Agiloft