API Reference: AI

The AI endpoints run AI prompts asynchronously and let you poll for the result.

Start a direct AI prompt (async; poll POST …/ai/{promptType}/status with the same body)

Endpoint: POST /v1/tenant/{tenantId}/ai/{promptType}

Auth: Tenant-scoped API key · write scope

Parameters:

Name	Type	Location	Required	Description
tenantId	string	path	Yes	Format: uuid
promptType	processAnalysis | processQuestion | aiSummary | processDocumentation | processRaci | processMetaEdit	path	Yes	Allowed values: processAnalysis, processQuestion, aiSummary, processDocumentation, processRaci, processMetaEdit

Request Body:

{
	"prompt": "string",
	"summaryVariant": "string"
}

Field	Type	Required	Description
prompt	string	Yes	Natural language prompt Max length: 10000
summaryVariant	string	No

Response (200):

AI result (status: success | generating | error)

{
	"status": "success",
	"result": {
		"type": "processAnalysis",
		"analysis": "string"
	},
	"error": "string"
}

Field	Type	Required	Description
status	success | error | generating	Yes	Allowed values: success, error, generating
result	object | object | object | object | object | object	No	Prompt-specific result payload (present when status is success)
error	string	No	Error message when status is error

Errors:

• 401 Missing, invalid, disabled or expired API key
• 429 Usage limit reached

Poll an in-flight AI generation (send the exact same body as the POST that started it)

POST-to-poll is deliberate: the generation is keyed by prompt + promptType + summaryVariant, so the poll repeats the body instead of addressing a run id.

Endpoint: POST /v1/tenant/{tenantId}/ai/{promptType}/status

Auth: Tenant-scoped API key · write scope

Parameters:

Name	Type	Location	Required	Description
tenantId	string	path	Yes	Format: uuid
promptType	processAnalysis | processQuestion | aiSummary | processDocumentation | processRaci | processMetaEdit	path	Yes	Allowed values: processAnalysis, processQuestion, aiSummary, processDocumentation, processRaci, processMetaEdit

Request Body:

{
	"prompt": "string",
	"summaryVariant": "string"
}

Field	Type	Required	Description
prompt	string	Yes	Natural language prompt Max length: 10000
summaryVariant	string	No

Response (200):

AI result (status: success | generating | error)

{
	"status": "success",
	"result": {
		"type": "processAnalysis",
		"analysis": "string"
	},
	"error": "string"
}

Field	Type	Required	Description
status	success | error | generating	Yes	Allowed values: success, error, generating
result	object | object | object | object | object | object	No	Prompt-specific result payload (present when status is success)
error	string	No	Error message when status is error

Errors:

• 401 Missing, invalid, disabled or expired API key
• 429 Usage limit reached