Workspace data API

All paths are relative to https://api.corpdev.ai/v1/public. The API key selects the workspace. The reference below identifies current route families; permissions and enabled capabilities still determine which operations your client can use.

Lists and pipelines

Method and path	Operation
GET /lists / POST /lists	List or create lists.
GET /lists/:id / PATCH /lists/:id	Read or update a list's supported settings.
POST /lists/:id/rows	Add rows.
PATCH /lists/:id/rows/:rowId / DELETE /lists/:id/rows/:rowId	Update or remove a row.
GET /pipelines / POST /pipelines	List or create pipelines.
GET /pipelines/:id/board	Read board information.
GET /pipelines/:id/deals	Read deals.
PATCH /pipelines/:id	Update supported pipeline fields.
POST /pipelines/:id/deals	Add a deal.
PATCH /pipelines/:id/deals/:dealId / DELETE /pipelines/:id/deals/:dealId	Update or remove a deal.
PATCH /pipelines/:id/config	Update supported pipeline configuration.

Use returned list, row, pipeline and deal identifiers; a company domain is not interchangeable with every row identifier. Changing deal stages can affect enabled automations. Confirm the target and schema before sending writes.

Investment criteria and overview

Investment target profiles use /itps: list and create at the collection, and read, patch or delete at /itps/:itpId. Company overview content is read or written at /companies/:domain/overview with GET or PUT.

Changing criteria can affect later fit assessments. Preserve the meaning and rationale of the criteria in the work product rather than treating a numeric score as independent evidence.

AI Room

Method and path	Operation
GET /ai-room/tree	Read the folder hierarchy.
GET /ai-room/folders/:folderId/contents	Read a folder's contents.
GET /ai-room/files/:fileId	Read an accessible file representation.
GET /ai-room/index-list	Request the room index list where supported.
POST /ai-room/search	Search source-file contents.
POST /ai-room/folders	Create a folder.
POST /ai-room/files	Upload using the route's multipart contract.
POST /ai-room/deep-research	Run supported research over room content.

An uploaded file may still need processing. Check its state before expecting extracted content. Keep file identifiers separate from deliverable document identifiers, and use the appropriate route family.

AI columns and trigger scans

For either lists/:containerId or pipelines/:containerId, append these paths:

Method and suffix	Operation
GET /ai-columns	List AI column configuration.
POST /ai-columns	Create an AI column.
PATCH or DELETE /ai-columns/:columnKey	Update or remove a column.
POST /ai-columns/:columnKey/populate	Start population.
POST /trigger-columns	Create a trigger column.
PATCH or DELETE /trigger-columns/:columnKey	Update or remove a trigger column.
POST /trigger-columns/:columnKey/scan	Start a scan.

For example, the full population route is POST /v1/public/lists/:containerId/ai-columns/:columnKey/populate. Read the column configuration first and use its returned key. This page is a route overview; configuration payloads depend on the column type and should be confirmed for your integration before writing. Use returned run identifiers with GET /ai-column-runs/:runId or GET /trigger-scan-runs/:runId. An AI-column run supports POST /ai-column-runs/:runId/cancel with write access.

A stop or cancel request should be checked against the resulting state; it does not erase work already completed. Batch analysis can consume credits.

Automations

Read automations with GET /automations and GET /automations/:id. POST /automations/plan prepares a plan. Create, update, delete, run and manifest-apply operations require write access and separate enablement for API automation changes.

Mutation paths include POST /automations, PATCH or DELETE /automations/:id, POST /automations/:id/run and POST /automations/apply. Review trigger scope, recipients and destination documents before applying a change. A run can produce a document or send configured messages.

Usage

GET /usage exposes usage information available to the key. Keep usage checks separate from rate-limit handling; a client can encounter either restriction independently.

See API behavior for errors, pagination, retries and metered reads.