Integration

V7 Darwin integration

V7 Darwin is a training-data platform for computer vision. Create and configure datasets, add images, videos, PDFs or DICOM files (by URL, base64 content or from connected cloud storage), list, count, move, prioritise, tag, archive and delete items in bulk with filters, drive them through annotation workflows (create and edit workflows, set stages, assign annotators), manage annotation classes, item properties, dataset labels and comment threads, export annotations (Darwin JSON, COCO, YOLO, Pascal VOC and more) and build annotator or item reports. External storage configuration, team membership changes and the three-step local file upload (signed URL) are not exposed. Authentication is an API key sent as 'ApiKey <key>' in the Authorization header; the connection also stores the team slug the key belongs to.

Connect V7 Darwin How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in V7 Darwin.Built from the 72 V7 Darwin tools below, on the connection you add once.

What you can do with V7 Darwin

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Label, 5 operations

create_dataset_labelCreate dataset labelPOSTCreate a dataset label (a colored tag for organising datasets), optionally attached to datasets.
delete_dataset_labelDelete dataset labelDELETEDelete a dataset label. The datasets themselves are not affected.
get_dataset_labelGet dataset labelGETGet one dataset label with its color and datasets.
list_dataset_labelsList dataset labelsGETList the team's dataset labels (colored tags used to organise datasets) with the datasets carrying each.
update_dataset_labelUpdate dataset labelPUTRename or recolor a dataset label, or change its datasets. Send only the fields to change.

Property, 5 operations

create_propertyCreate propertyPOSTCreate a property: a custom field annotators fill per item, annotation or section (single_select, multi_select, boolean, text or audio), with its values.
delete_propertyDelete propertyDELETEDelete a property together with all its values.
get_propertyGet propertyGETGet one property with its values, type, granularity and dataset scope.
list_propertiesList propertiesGETList the team's properties (custom fields annotators fill: single or multi select, boolean, text, audio) with their values.
update_propertyUpdate propertyPUTUpdate a property and add values to it. The name is required by the vendor's update schema: read get_property first and send it back with the change.

Workflow, 5 operations

archive_workflowArchive workflowDELETEArchive a workflow. Returns the archived workflow.
create_workflowCreate workflowPOSTCreate an annotation workflow from a list of stages (dataset, annotate, review, complete ...). Copy the stage shape from get_workflow on an existing workflow.
get_workflowGet workflowGETGet one workflow with its stages (ids, types, assignees, edges), linked dataset and progress.
list_workflowsList workflowsGETList the team's annotation workflows with their stages, linked dataset and progress.
update_workflowUpdate workflowPATCHRename a workflow or change its stages. stages REPLACES the stage list: read get_workflow first and send the full list back with your change.

Annotation class, 4 operations

create_annotation_classCreate annotation classPOSTCreate an annotation class in the team, with one main annotation type (for example polygon, bounding_box, tag) and optional subtypes, optionally attached to datasets.
delete_annotation_classDelete annotation classDELETEPermanently delete an annotation class AND every annotation that uses it. Cannot be undone.
list_annotation_classesList annotation classesGETList the team's annotation classes (name, annotation types such as polygon or tag, color metadata, dataset) and counts per annotation type. Use a class id in item filters or tag_items.
update_annotation_classUpdate annotation classPUTUpdate an annotation class: name, description, metadata, annotation types or datasets. datasets and annotation_type_ids REPLACE the current associations.

Comment, 4 operations

create_commentCreate commentPOSTReply to a comment thread.
delete_commentDelete commentDELETEDelete a comment. Only its author can.
list_commentsList commentsGETList the comments of a comment thread.
update_commentUpdate commentPATCHEdit the text of a comment. Only its author can.

Comment thread, 4 operations

create_comment_threadCreate comment threadPOSTStart a comment thread at a position on one slot of an item, with its first comment.
delete_comment_threadDelete comment threadDELETEDelete a comment thread and its comments.
list_comment_threadsList comment threadsGETList the comment threads on an item, with position, resolved flag, first comment and comment count.
update_comment_threadUpdate comment threadPATCHResolve or reopen a comment thread, or move its bounding box.

Dataset, 4 operations

create_datasetCreate datasetPOSTCreate an empty dataset in the team. Returns its id and slug; add items with upload_items or register_storage_items.
get_datasetGet datasetGETGet one dataset by id: name, slug, settings, item and class counts, instructions and its annotation classes.
list_datasetsList datasetsGETList the team's datasets with id, slug, item counts, annotation progress and annotator settings. Use a dataset id in item filters and its slug for exports and imports.
update_datasetUpdate datasetPUTChange a dataset's name, annotator instructions and annotation settings. Send only the fields to change.

Export, 4 operations

create_exportCreate exportPOSTStart an export (release) of a dataset's annotations in a chosen format. Processed asynchronously: poll get_export until its status is complete, then read download_url.
delete_exportDelete exportDELETEDelete an export and its stored files. A running export job is not cancelled.
get_exportGet exportGETGet one export of a dataset by name: status, format, version and download_url once complete.
list_exportsList exportsGETList a dataset's exports (releases), including pending ones, with format, status, version and download_url.

Item, 4 operations

delete_itemsDelete itemsDELETEPERMANENTLY delete every item matched by filters, with their annotations. Cannot be undone: list_items with the same filters first to check the selection.
get_itemGet itemGETGet one item: name, path, dataset, status, priority, layout, tags and its slots (files) with their URLs and metadata.
list_itemsList itemsGETList items (files) of one or more datasets, filtered by name, path, status, type, stage, class or assignee. Pages of up to 500; pass page[from] with the returned next cursor.
rename_itemRename itemPUTRename an item. The name must not contain / and must be unique in its folder and dataset; a deleted item cannot be updated.

Job, 4 operations

create_report_jobCreate report jobPOSTStart a background job building an annotator report (data from 2024-11-05) or item report (from 2025-08-01) as CSV or JSON. Poll get_report_job for its status and file URL.
delete_report_jobDelete report jobDELETEDelete a report job.
get_report_jobGet report jobGETGet one report job: status and the URL of the report file once generated.
list_report_jobsList report jobsGETList the team's report jobs with status and file URL, optionally of one type or dataset.

Archive, 2 operations

archive_datasetArchive datasetPUTArchive a dataset (it disappears from the active list). Returns the archived dataset.
archive_itemsArchive itemsPOSTArchive every item matched by filters (restore_items brings them back). Returns the number of items affected.

Property value, 2 operations

delete_property_valueDelete property valueDELETEDelete one value of a property.
update_property_valueUpdate property valuePUTChange the text, color or position of one value of a property.

Tag, 2 operations

tag_itemsTag itemsPOSTAdd a tag annotation (an annotation class of type tag) to every slot of the items matched by filters.
untag_itemsUntag itemsDELETERemove a tag annotation from every slot of the items matched by filters.

Annotator, 1 operation

get_annotator_reportGet annotator reportGETBuild an annotator report for a period (data from 2024-11-05): metrics such as active_time or total_annotations, grouped by annotators, datasets, stages and more.

Assign, 1 operation

assign_itemsAssign itemsPOSTAssign an annotator (by assignee_id or assignee_email) to every item matched by filters in a workflow.

Batch, 1 operation

update_dataset_labelsUpdate dataset labelsPOSTSet the datasets of several dataset labels in one call. Each entry names a label id and its dataset ids.

Class count, 1 operation

count_items_by_classCount items by classGETCount the items matching the filters per annotation class.

Direct upload, 1 operation

upload_itemsUpload itemsPOSTAdd items to a dataset from public file URLs or base64 file content in one call. Returns the created items and any blocked ones with a reason (for example ALREADY_EXISTS).

Folder, 1 operation

list_item_foldersList item foldersGETList the folders of one or more datasets, with item counts, optionally only those under current_path or only non-empty ones.

General count, 1 operation

count_itemsCount itemsGETCount the items matching the filters per dataset; the answer also gives the counts with no filter applied.

Layout, 1 operation

set_item_layoutSet item layoutPOSTSet how the slots (files) of the matched multi-file items are arranged in the annotation view.

List id, 1 operation

list_item_idsList item idsGETList only the ids of the items matching the filters, paged like list_items. Useful before a bulk operation on item_ids.

Membership, 1 operation

list_team_membersList team membersGETList the members of the team the API key belongs to: user id, name, email and role.

Path, 1 operation

move_itemsMove itemsPOSTMove every item matched by filters to a folder path (absolute, for example /batch-2). Returns the number of items affected.

Priority, 1 operation

set_item_prioritySet item priorityPOSTSet the annotation priority of every item matched by filters. Returns the number of items affected.

Register existing, 1 operation

register_storage_itemsRegister storage itemsPOSTRegister files already in the team's connected cloud storage (read-write: Darwin writes thumbnails back to the storage). Name the storage by its slug and each file by storage_key.

Register existing readonly, 1 operation

register_storage_items_readonlyRegister storage items readonlyPOSTRegister files in the team's connected cloud storage read-only: Darwin does not process them or write thumbnails, so thumbnail paths must be given per slot. Rate-limited to 100 requests per minute.

Request work batch, 1 operation

request_work_batchRequest work batchPOSTRequest a batch of items to work on in a workflow for the API key's user, as an annotator would. Returns the ids of the items assigned.

Response, 1 operation

list_webhook_responsesList webhook responsesGETList the latest responses (20 successful, 20 failed) a webhook stage's URL returned for items passing through it, with response code and body.

Restore, 1 operation

restore_itemsRestore itemsPOSTRestore archived items matched by filters. Returns the number of items affected.

Section, 1 operation

list_slot_sectionsList slot sectionsGETList the sections (frames or pages) of one slot (file) of an item, with their URLs. Page from a given index with page[from].

Stage, 1 operation

set_item_stageSet item stagePOSTMove every item matched by filters to a stage of a workflow.

Stage count, 1 operation

count_items_by_stageCount items by stageGETCount the items matching the filters per current workflow stage.

Status count, 1 operation

count_items_by_statusCount items by statusGETCount the items matching the filters per status (new, annotate, review, complete, archived ...).

Time summary, 1 operation

get_item_time_summaryGet item time summaryGETGet the time spent on an item, in its current workflow and outside it.

Unlink dataset, 1 operation

unlink_workflow_datasetUnlink workflow datasetPATCHDetach the dataset from a workflow's dataset stage.

Use V7 Darwin from Claude Code, Cursor or Codex

Connect V7 Darwin in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its V7 Darwin operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect V7 Darwin

Create your workspace
Sign up free. Nothing to install.
Connect V7 Darwin once
Add your V7 Darwin connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.