Debug Worklogs API

The Debug Worklogs API provides detailed debug log data for workflow runs, including AI Employee configuration snapshots, workflow definitions, and per-action show-work logs. Access is filtered based on the requesting user's permission level.

GetWorkflowLevelDebugLog

Retrieve comprehensive debug data for an entire workflow run.

Property	Value
URL	`/persona.v1.DebugLogService/GetWorkflowLevelDebugLog`
HTTP Method	POST
Protocol	gRPC-Web over HTTP

Headers

Header	Type	Required	Description
`Authorization`	string	Yes	Bearer token
`Content-Type`	string	Yes	`application/grpc-web+proto`
`x-grpc-web`	string	Yes	`1`

Request

Field	Type	Required	Description
`workflow_run_id`	string	Yes	UUID of the workflow run to debug

Response

Field	Type	Description
`persona_config`	PersonaConfigForShowWork	AI Employee configuration snapshot (Debug users only)
`action_types`	map<string, ActionType>	Map of action names to their type snapshots
`workflow_definition`	WorkflowDef	Workflow definition snapshot at time of execution (Debug users only)
`show_work_logs`	map<string, ShowWorkLog>	Map of action instance names to their show-work logs
`is_ema_user`	bool	Whether the requester is an Ema user

Example Response

{
 "persona_config": {
 "id": "persona-uuid",
 "name": "Support Agent",
 "description": "Customer support persona",
 "status": "ACTIVE",
 "version": 1,
 "widgets": [],
 "display_settings": {},
 "project_settings": {}
 },
 "action_types": {
 "search@v2": {
 "type_name": {},
 "display_name": "Knowledge Search",
 "description": "Searches the knowledge base"
 }
 },
 "workflow_definition": {
 "workflow_name": {},
 "actions": [],
 "display_name": "Support Workflow"
 },
 "show_work_logs": {
 "search_step": {
 "action_title": "Knowledge Search",
 "llm_call_count": 2,
 "llm_cost_usd": 0.05,
 "action_run_status": "ACTION_RUN_STATUS_SUCCESS"
 }
 },
 "is_ema_user": false
}

Permission Levels

The API filters response data based on the requesting user's access level:

User Type	Sees
Debug access user	Everything
Standard user	Action names and step names only

Data Structures

ShowWorkLog

Per-action debug information for a workflow run.

Field	Type	Description
`messages`	repeated StructuredLogMessage	Log messages with substitutable variables
`llm_call_count`	int32	Number of LLM calls made
`action_title`	string	Display title of the action
`llm_calls`	repeated LlmCallEntry	Detailed LLM call entries
`llm_cost_usd`	float	Total LLM cost in USD
`llm_latency_ms`	float	Total LLM latency in milliseconds
`inputs`	ShowWorkInputs	Detailed input information
`outputs`	ShowWorkOutputs	Detailed output information
`hitl_rounds`	repeated HitlRound	Human-in-the-loop interaction rounds
`work_logs`	repeated Step	Steps executed in the action
`workflow_run_id`	string	UUID of the workflow run
`action_run_status`	ActionRunStatus	Status of the action run
`embedded_persona_run_info`	EmbeddedPersonaRunInfo	Info about embedded AI Employee runs

ActionRunStatus Enum

Value	Integer	Description
`ACTION_RUN_STATUS_UNSPECIFIED`	0	Status not specified
`ACTION_RUN_STATUS_SUCCESS`	1	Action ran successfully
`ACTION_RUN_STATUS_ERRORED`	2	Action encountered an error
`ACTION_RUN_STATUS_WARNING`	3	Action completed with warnings
`ACTION_RUN_STATUS_NOT_RUN`	4	Action was not executed
`ACTION_RUN_STATUS_PAUSED`	5	Action execution was paused

StructuredLogMessage

Field	Type	Description
`message_template`	string	Template with `{placeholder}` variables
`substituted_fields`	map<string, string>	Placeholder-to-value mappings
`heading`	string	Optional heading
`log_level`	LogLevel	Message severity level
`error_info`	ErrorInfo	Error details (if applicable)

LlmCallEntry

Field	Type	Description
`metrics`	ResponseMetrics	LLM response metrics
`fusion_used`	bool	Whether Fusion selected a model
`success`	bool	Whether the LLM call succeeded
`error_message`	string	Error message if failed
`timestamp_ms`	int64	Timestamp in milliseconds
`llm_cost_final_choice`	float	Cost of the final model choice in USD
`prompt`	PromptData	Prompt data (super-admin only)
`heading`	string	Display heading for the call

PromptData

Field	Type	Description
`system_prompt`	string	System prompt text
`user_prompt`	string	User prompt text
`response`	string	LLM response text

HitlRound

Field	Type	Description
`message_type`	MessageType	Type of HITL message (Form, Button, TextField, Feedback)
`hitl_request`	oneof	Message shown to the human
`hitl_response`	oneof	Response from the human
`outcome`	string	Outcome status of the round

ShowWorkInputs / ShowWorkOutputs

Field	Type	Description
`entries`	repeated InputEntry/OutputEntry	List of input/output entries

InputEntry

Field	Type	Description
`name`	string	Parameter name
`value`	Value	The actual value
`type`	ResolvedType	Data type
`source_agent_name`	string	Name of the source action

OutputEntry

Field	Type	Description
`name`	string	Parameter name
`value`	Value	The actual value
`type`	ResolvedType	Data type
`published_as_chat_response`	bool	Whether published to the end user

ErrorInfo

Field	Type	Description
`name`	string	Error name
`internal_message`	string	Internal error message (not shown to end users)
`external_message`	string	Error message shown to end users

Step

Field	Type	Description
`name`	string	Step name
`messages`	repeated StructuredLogMessage	Log messages for the step
`llm_calls`	repeated LlmCallEntry	LLM calls in the step

Error Responses

HTTP Status	gRPC Code	Description
400	`INVALID_ARGUMENT`	`workflow_run_id` is required or invalid UUID
403	`PERMISSION_DENIED`	User lacks permission for this AI Employee
404	`NOT_FOUND`	Workflow run not found or tenant mismatch
500	`INTERNAL`	Internal server error

Workflow Overview -- Workflow concepts
Action Data Model -- ActionType structure
Handling gRPC-Web Requests -- gRPC-Web encoding guide