Harvey
App · 20 actions
Work with legal intake requests, Vault projects and Harvey Assistant.
Actions
| Action | What it does |
|---|---|
| Add review table rows | Adds one or more files to an existing review table as new rows and populates each new row's cells for every existing visible column. |
| Ask Harvey Assistant | Harvey's Completion API provides access to Harvey's models via our orchestration engine, allowing users to ask complex, freeform questions and receive detailed legal-grade responses programmatically. |
| Create intake request | Submits a request into your workspace's Intake queue and returns its identifier. |
| Create Vault project | Creates a new project in the Vault system. |
| Delete Vault project | Delete Vault project. |
| Get file details | Retrieves details and processing status for one or more files by their IDs. |
| Get intake request | Fetches the status for an intake request. |
| Get query history | Fetches the query history and metadata for a specified time range, including sensitive information such as inputs to and outputs from the AI system. |
| Get review table | Fetches metadata for a single review table by its ID, including its title, creation timestamp, and the list of file IDs associated with it. |
| Get review table row | Enables authorized users to retrieve a row from a review table, given a review table ID and a file ID corresponding to the row. |
| Get usage history | Fetches the usage history and metadata for a specified time range. |
| Get Vault project | Fetches metadata for a specified Vault project, including file details and storage limits. |
| List deleted Vault projects | Lists every vault currently in the recycle bin for the current workspace. |
| List intake comments | Lists the comments on an intake request, newest first. |
| List intake requests | Lists the intake requests your token's user can see, newest first. |
| List Vault files | Lists files within a single vault project, with filtering and sorting supported. |
| List Vault project users | Lists all workspace users with current active access to a specified Vault project, along with their access levels. |
| List Vault projects | Lists all Vault projects and knowledge bases in the current workspace with pagination support. |
| Search Vault documents | Performs semantic search over documents in a single vault project. |
| Update intake request | Changes the fields sent in the body and leaves every other field as it was. |
Add review table rows
Adds one or more files to an existing review table as new rows and populates each new row's cells for every existing visible column.
Inputs
- Review table ID
review_table_idnumberRequired - The review table ID (integer). This is the same value as review_table_ids entries returned by Get Vault Project Metadata and as review_table_id returned by Get Review Table Metadata.
- File IDS
file_idslistRequired - The IDs of the files to add. One file is added as its own row; two or more are grouped into a file group that occupies a single row. Every file must belong to the review table's Vault project, must have finished processing, and must not already have a row or group membership…
- Group name
group_nametext - Optional name for the file group created when two or more files are added. Defaults to the first file's name. Ignored for single-file requests.
Outputs
dataobject- Result.
Example output
{
"data": {
"response": {
"event_id": 1,
"file_ids": [],
"review_event_run_id": "example",
"review_file_group_id": "example",
"status": "example"
}
}
}Ask Harvey Assistant
Harvey's Completion API provides access to Harvey's models via our orchestration engine, allowing users to ask complex, freeform questions and receive detailed legal-grade responses programmatically. It can also be used in conjunction with files attached to the request,…
Inputs
- Include citations
include_citationstrue/false - Whether to include citations in the response output. When set to false, citations are skipped and the response is returned more quickly. Defaults to true.Default:
true. - Prompt
prompttextRequired - A string question that you want to ask Harvey. The prompt may be up to 20,000 characters.
- Stream
streamtrue/false - A boolean value that lets you either stream or receive the entire output at once.Default:
false. - File
filelist - Files you want to upload to provide additional context for the prompt.
- Mode
modetext - Optional mode to specify the behavior of the completion.One of:
draft,assist. Default:draft. - Client matter ID
client_matter_idtext - Optional client matter identifier to associate with this completion.
- Knowledge sources
knowledge_sourcestext - JSON-encoded array of knowledge sources such as Vaults or Regional Knowledge Sources.
- Model
modeltext - Optional model selector. Pass a customer-facing model slug to use a specific model for this request. Omit this parameter to use your workspace's default model.One of:
gpt-5,gpt-5-1,gpt-5-2,gpt-5-4,gpt-5-4-mini,gpt-5-5,gpt-4-1,claude-opus-4-5,claude-sonnet-4-5,claude-opus-4-6, and 5 more.
Outputs
dataobject- Result.
Example output
{
"data": {
"response": "example",
"response_with_citations": "example",
"sources": []
}
}Create intake request
Submits a request into your workspace's Intake queue and returns its identifier. Read the intake request for its progress.
Inputs
- Idempotency key
idempotency_keytextRequired - An identifier you choose for this request, unique within your workspace. A retry carrying a key Harvey has already accepted returns the original request with 200 instead of creating a second one. Must not be blank or whitespace only.
- Title
titletextRequired - Short summary of what is being asked for. Surrounding whitespace is stripped, and the result must not be empty.
- Description
descriptiontext - The full request, in as much detail as the reviewer needs.
- Files
fileslist - A document the request concerns. Repeat the field once per document. Up to 50 files, and up to 100 MB per file for most types. A file whose type or size is not accepted refuses the entire request, so a request never silently loses a document.
Outputs
dataobject- Returned when a request is submitted, and when a retry replays an idempotency key Harvey already accepted.
Example output
{
"data": {
"harvey_url": "example",
"request_id": "example",
"status": "example"
}
}Create Vault project
Creates a new project in the Vault system. The project can be designated as a knowledge base.
Inputs
- Name
nametextRequired - The name of the new project
- Is knowledge base project
is_knowledge_base_projecttrue/false - Flag indicating whether this project is a knowledge base
- Client matter ID
client_matter_idtext - When set, associates the new project with this client matter. Use the client matter UUID (from Client Matters API responses), not the human-readable matter number.
Outputs
dataobject- Result.
Example output
{
"data": {
"created_at": "example",
"description": "example",
"name": "example",
"project_id": "example",
"settings": {
"allowed_file_types": [],
"max_file_size": 1,
"max_files": 1
},
"status": "example"
}
}Delete Vault project
Delete Vault project.
Inputs
- Project
project_idtextRequired - The unique identifier of the project to delete
Outputs
dataobject- Result.
Example output
{
"data": {
"message": "example"
}
}Get file details
Retrieves details and processing status for one or more files by their IDs. A maximum of 1000 file IDs can be provided per request. This endpoint is useful for checking the processing status of recently uploaded files to determine when they are ready to query. Files that are…
Inputs
- File IDS
file_idslistRequired - Comma-separated list of file IDs to retrieve details for.
Outputs
dataobject- Result.
Example output
{
"data": {
"response": {
"content": []
}
}
}Get intake request
Fetches the status for an intake request.
Inputs
- Request
request_idtextRequired - The unique identifier of the intake request, as returned when it was submitted.
Outputs
dataobject- Result.
Example output
{
"data": {
"assignee": {
"name": "example",
"type": "example"
},
"attachments": [],
"completed_at": "example",
"created_at": "example",
"description": "example",
"due_at": "example",
"harvey_url": "example",
"request_id": "example",
"status": "example",
"title": "example",
"updated_at": "example"
}
}Get query history
Fetches the query history and metadata for a specified time range, including sensitive information such as inputs to and outputs from the AI system. Returns enriched event data with product surface area, subsurface details, and additional metadata.
Inputs
- Start time
start_timedate and timeRequired - UTC epoch timestamp, up to 1 year ago from now
- End time
end_timedate and timeRequired - UTC epoch timestamp, up to 30 days from start_time. Note: The API supports a maximum range of 30 days, but it's best to retrieve a max of one day at a time on a regular cadence.
- Include file IDS
include_file_idstrue/false - When set to true, the response will include file_ids for each event. This may impact performance, so only enable when needed (e.g., to use with the file download endpoint).Default:
false.
Outputs
dataobject- Result.
Example output
{
"data": {
"events": []
}
}Get review table
Fetches metadata for a single review table by its ID, including its title, creation timestamp, and the list of file IDs associated with it.
Inputs
- Review table ID
review_table_idnumberRequired - The integer ID of the review table. Obtain this from the review_table_ids array in the Get Vault Project Metadata response.
Outputs
dataobject- Result.
Example output
{
"data": {
"response": {
"content": {
"created_at": "example",
"file_ids": [],
"review_table_id": 1,
"title": "example"
}
}
}
}Get review table row
Enables authorized users to retrieve a row from a review table, given a review table ID and a file ID corresponding to the row.
Inputs
- Review table ID
review_table_idnumberRequired - The review table ID (integer). This is the same value as review_table_ids entries returned by Get Vault Project Metadata and as review_table_id returned by Get Review Table Metadata.
- File ID
file_idtextRequired - The file ID corresponding to the row. A file that is part of a file group resolves to the group's row.
Outputs
dataobject- Result.
Example output
{
"data": {
"assigned_user": "example",
"event_id": 1,
"file_id": "example",
"is_assigned": false,
"response": {
"cells": []
},
"updated_at": "example"
}
}Get usage history
Fetches the usage history and metadata for a specified time range. Returns enriched event data with product surface area, subsurface details, and additional metadata across Harvey product areas.
Inputs
- Start time
start_timedate and timeRequired - UTC epoch timestamp, up to 1 year ago from now
- End time
end_timedate and timeRequired - UTC epoch timestamp, up to 30 days from start_time. Note: The API supports a maximum range of 30 days, but it's best to retrieve a max of one day at a time on a regular cadence.
- Include file IDS
include_file_idstrue/false - When set to true, the response will include file_ids for each event. This may impact performance, so only enable when needed (e.g., to use with the file download endpoint).Default:
false.
Outputs
dataobject- Result.
Example output
{
"data": {
"events": []
}
}Get Vault project
Fetches metadata for a specified Vault project, including file details and storage limits.
Inputs
- Project
project_idtextRequired - The project ID.
Outputs
dataobject- Result.
Example output
{
"data": {
"response": {
"content": []
}
}
}List deleted Vault projects
Lists every vault currently in the recycle bin for the current workspace. When a Vault project is deleted it is not purged immediately. it first moves to a workspace recycle bin, where it can be reviewed before being permanently deleted. Entries are ordered by deleted_at…
Inputs
- Search text
search_texttext - Case-insensitive substring filter on the vault name.
- Client matter ID
client_matter_idtext - Restrict results to vaults associated with this client matter (the client matter UUID, not the human-readable matter number).
Outputs
dataobject- Result.
Example output
{
"data": {
"items": [],
"total": 1
}
}List intake comments
Lists the comments on an intake request, newest first. Follow pagination.next_cursor until pagination.has_more is false.
Inputs
- Request
request_idtextRequired - The unique identifier of the intake request, as returned when it was submitted.
- Limit
limitnumber - The page size. Values outside the range are clamped rather than rejected.From 1 to 100. Default:
20. - Cursor
cursortext - The pagination.next_cursor from the previous page of this request's comments.
Outputs
dataobject- Result.
Example output
{
"data": {
"comments": [],
"pagination": {
"has_more": false,
"limit": 1,
"next_cursor": "example"
}
}
}List intake requests
Lists the intake requests your token's user can see, newest first. Follow pagination.next_cursor until pagination.has_more is false.
Inputs
- Limit
limitnumber - The page size. Values outside the range are clamped rather than rejected.From 1 to 100. Default:
20. - Cursor
cursortext - The pagination.next_cursor from the previous page. Send the same filters with it, since a cursor resumes only the list it was issued for.
- Status
statuslist - Returns only requests in one of these statuses. Repeat it for several, as in status=TODO&status=IN_PROGRESS. No other parameter may repeat.One or more of:
TODO,IN_PROGRESS(In progress),READY_FOR_REVIEW(Ready for review),DONE,CANCELLED. - Created after
created_afterdate and time - Returns only requests created at or after this time. The value must include a UTC offset.
- Created before
created_beforedate and time - Returns only requests created before this time. The value must include a UTC offset.
Outputs
dataobject- Result.
Example output
{
"data": {
"pagination": {
"has_more": false,
"limit": 1,
"next_cursor": "example"
},
"requests": []
}
}List Vault files
Lists files within a single vault project, with filtering and sorting supported.
Inputs
- Project
project_idtextRequired - The unique identifier of the vault project
- Cursor
cursortext - Opaque pagination cursor returned in the previous response's next_cursor field. Omit on the first request.
- Limit
limitnumber - Maximum number of files to return per page.From 1 to 100. Default:
20. - Name
nametext - Case-insensitive substring match on file name. % and _ are treated as literals, not SQL LIKE wildcards.
- Content type
content_typelist - Filter by MIME type. Repeat the query parameter to match any of multiple values (e.g. ?content_type=application/pdf&content_type=text/plain). Up to 20 values, each at most 256 characters.
- Processing status
processing_statuslist - Filter by processing status. Repeat the query parameter to match any of multiple values. Accepted case-insensitively (READY_TO_QUERY and ready_to_query are equivalent). Internal-only statuses are rejected. Up to 20 values.One or more of:
UPLOADED,PROCESSING,READY_TO_QUERY(Ready to query),READY_TO_REVIEW(Ready to review),RECOVERABLE_FAILURE(Recoverable failure),UNRECOVERABLE_FAILURE(Unrecoverable failure). - Uploaded after
uploaded_afterdate and time - Inclusive lower bound on the file's upload time, as a UTC Unix epoch timestamp in seconds. Must be strictly less than uploaded_before when both are provided.
- Uploaded before
uploaded_beforedate and time - Inclusive upper bound on the file's upload time, as a UTC Unix epoch timestamp in seconds.
- Sort by
sort_bytext - Field to sort the results by. Accepted case-insensitively.One of:
name,uploaded_at(Uploaded at),size. Default:uploaded_at. - Sort order
sort_ordertext - Sort direction. Accepted case-insensitively. Results include the file id as a deterministic tiebreaker so cursor-paginated ordering is stable across pages.One of:
asc,desc. Default:desc.
Outputs
dataobject- Result.
Example output
{
"data": {
"response": {
"content": {
"files": [],
"pagination": {
"has_more": false,
"next_cursor": "example",
"page_size": 1,
"total_count": 1
}
}
}
}
}List Vault project users
Lists all workspace users with current active access to a specified Vault project, along with their access levels. Includes effective user access (including inherited access), and the project owner is always included with MANAGE access.
Inputs
- Project
project_idtextRequired - The unique identifier of the Vault project
Outputs
dataobject- Result.
Example output
{
"data": {
"data": {
"project_id": "example",
"users": []
},
"status": "example"
}
}List Vault projects
Lists all Vault projects and knowledge bases in the current workspace with pagination support. This endpoint returns a paginated list of projects that the authenticated user has permission to view.
Inputs
- Page
pagenumber - Page number for pagination (default: 1)At least 1. Default:
1. - Per page
per_pagenumber - Number of items per page (default: 20, max: 100)From 1 to 100. Default:
20. - Name
nametext - Case-insensitive substring match on project name.
- Is knowledge base
is_knowledge_basetrue/false - Filter by knowledge base flag. true returns only knowledge base projects, false excludes them, and omitting the parameter returns both.
- Created after
created_afterdate and time - Inclusive lower bound on project creation time, expressed as a UTC Unix epoch timestamp in seconds. Only projects with created_at >= created_after are returned. Must be strictly less than created_before when both are provided.
- Created before
created_beforedate and time - Exclusive upper bound on project creation time, expressed as a UTC Unix epoch timestamp in seconds. Only projects with created_at < created_before are returned.
- Sort by
sort_bytext - Field to sort the results by. date sorts by the project's last content-update timestamp (the default ordering). name sorts alphabetically by project name.One of:
date,name. Default:date. - Sort order
sort_ordertext - Sort direction to order results list.One of:
asc,desc. Default:desc.
Outputs
dataobject- Result.
Example output
{
"data": {
"response": {
"content": {
"pagination": {
"page": 1,
"per_page": 1,
"total": 1,
"total_pages": 1
},
"projects": []
}
}
}
}Search Vault documents
Performs semantic search over documents in a single vault project. Returns the most relevant text chunks for the given natural-language query. Intended for use by external systems (e.g. a customer's own assistant or LLM) that need to retrieve relevant passages and then…
Inputs
- Project
project_idtextRequired - UUID of the vault project to search in
- Query
querytextRequired - Natural language search query.
- Include file names
include_file_namestrue/false - When true, each chunk in the response includes a file_name field (display name of the source file, or null if the file was deleted after indexing). Omit or set to false for backward compatibility.Default:
false.
Outputs
dataobject- Response from semantic search
Example output
{
"data": {
"data": {
"chunks": [],
"file_ids": []
},
"status": "example"
}
}Update intake request
Changes the fields sent in the body and leaves every other field as it was.
Inputs
- Request
request_idtextRequired - The unique identifier of the intake request, as returned when it was submitted.
- Status
statustext - Lifecycle state of an intake request.One of:
TODO,IN_PROGRESS(In progress),READY_FOR_REVIEW(Ready for review),DONE,CANCELLED. - Due at
due_atdate - The new due date as YYYY-MM-DD. Send null to remove it.
Outputs
dataobject- Result.
Example output
{
"data": {
"assignee": {
"name": "example",
"type": "example"
},
"attachments": [],
"completed_at": "example",
"created_at": "example",
"description": "example",
"due_at": "example",
"harvey_url": "example",
"request_id": "example",
"status": "example",
"title": "example",
"updated_at": "example"
}
}