Skip to content

Interface: AgentTraceStep

@kortexya/reasoninglayer


@kortexya/reasoninglayer / AgentTraceStep

Defined in: src/types/verification.ts:273

One step of an agent’s execution trajectory.

input and output are arbitrary JSON, sent verbatim.

action: string

Defined in: src/types/verification.ts:277

Tool or action name.


optional agentId?: string

Defined in: src/types/verification.ts:291

Id of the agent that executed the step, in a multi-agent system.


optional agentName?: string

Defined in: src/types/verification.ts:293

Name of the agent that executed the step, in a multi-agent system.


optional error?: string

Defined in: src/types/verification.ts:287

Error message, when the step failed.


optional fileContent?: string

Defined in: src/types/verification.ts:295

Content a file-reading tool returned.


input: JsonValue

Defined in: src/types/verification.ts:279

Tool input parameters (arbitrary JSON).


output: JsonValue

Defined in: src/types/verification.ts:281

Tool output or result (arbitrary JSON).


stepIndex: number

Defined in: src/types/verification.ts:275

Step index in the trajectory.


optional success?: boolean

Defined in: src/types/verification.ts:285

Whether the tool succeeded.


optional timestamp?: string

Defined in: src/types/verification.ts:289

ISO 8601 timestamp of the step.


optional toolUsed?: string

Defined in: src/types/verification.ts:283

Specific tool identifier.