OSS-first docs

These docs teach the open system first: contracts, generated surfaces, runtimes, governance, and incremental adoption. Studio shows up as the operating layer on top, not as the source of truth.

agent.run.execute

Starts a new agent run with the given input.

  • Type: operation (command)
  • Version: 1.0.0
  • Stability: stable
  • Owners: @agent-console-team
  • Tags: run, execute
  • File: packages/examples/agent-console/src/run/run.operation.ts
  • field.key.label
    agent.run.execute
    field.version.label
    1.0.0
    field.type.label
    operation (command)
    field.title.label
    agent.run.execute
    field.description.label

    Starts a new agent run with the given input.

  • Type: operation (command)
  • Version: 1.0.0
  • Stability: stable
  • Owners: @agent-console-team
  • Tags: run, execute
  • File: packages/examples/agent-console/src/run/run.operation.ts
  • field.tags.label
    run,execute
    field.owners.label
    @agent-console-team
    field.stability.label
    stable

    Starts a new agent run with the given input.

    Goal

    Execute an AI agent with user input.

    Context

    Called from chat interface or API.

    Source Definition

    import {
    	defineCommand,
    	defineQuery,
    } from '@lssm-tech/lib.contracts-spec/operations';
    import { defineSchemaModel, ScalarTypeEnum } from '@lssm-tech/lib.schema';
    import { GranularityEnum, LogLevelEnum, RunStatusEnum } from './run.enum';
    import {
    	RunInputModel,
    	RunLogModel,
    	RunModel,
    	RunStepModel,
    	RunSummaryModel,
    	TimelineDataPointModel,
    } from './run.schema';
    
    export const ExecuteAgentCommand = defineCommand({
    	meta: {
    		key: 'agent.run.execute',
    		version: '1.0.0',
    		stability: 'stable',
    		owners: ['@agent-console-team'],
    		tags: ['run', 'execute'],
    		description: 'Starts a new agent run with the given input.',
    		goal: 'Execute an AI agent with user input.',
    		context: 'Called from chat interface or API.',
    	},
    	io: {
    		input: defineSchemaModel({
    			name: 'ExecuteAgentInput',
    			fields: {
    				agentId: { type: ScalarTypeEnum.String_unsecure(), isOptional: false },
    				input: { type: RunInputModel, isOptional: false },
    				sessionId: { type: ScalarTypeEnum.String_unsecure(), isOptional: true },
    				metadata: { type: ScalarTypeEnum.JSONObject(), isOptional: true },
    				stream: { type: ScalarTypeEnum.Boolean(), isOptional: true },
    				maxIterations: {
    					type: ScalarTypeEnum.Int_unsecure(),
    					isOptional: true,
    				},
    				timeoutMs: { type: ScalarTypeEnum.Int_unsecure(), isOptional: true },
    			},
    		}),
    		output: defineSchemaModel({
    			name: 'ExecuteAgentOutput',
    			fields: {
    				runId: { type: ScalarTypeEnum.String_unsecure(), isOptional: false },
    				status: { type: RunStatusEnum, isOptional: false },
    				estimatedWaitMs: {
    					type: ScalarTypeEnum.Int_unsecure(),
    					isOptional: true,
    				},
    			},
    		}),
    		errors: {
    			AGENT_NOT_FOUND: {
    				description: 'The specified agent does not exist',
    				http: 404,
    				gqlCode: 'AGENT_NOT_FOUND',
    				when: 'Agent ID is invalid',
    			},
    			AGENT_NOT_ACTIVE: {
    				description: 'The specified agent is not active',
    				http: 400,
    				gqlCode: 'AGENT_NOT_ACTIVE',
    				when: 'Agent is in draft/paused/archived state',
    			},
    		},
    	},
    	policy: { auth: 'user' },
    	sideEffects: {
    		emits: [
    			{
    				key: 'run.started',
    				version: '1.0.0',
    				stability: 'stable',
    				owners: ['@agent-console-team'],
    				tags: ['run', 'started'],
    				when: 'Run is queued',
    				payload: RunSummaryModel,
    			},
    		],
    		audit: ['run.started'],
    	},
    	acceptance: {
    		scenarios: [
    			{
    				key: 'execute-agent-happy-path',
    				given: ['Agent exists', 'Agent is active'],
    				when: ['User submits execution request'],
    				then: ['Run is created', 'RunStarted event is emitted'],
    			},
    			{
    				key: 'execute-agent-not-active',
    				given: ['Agent exists but is not active'],
    				when: ['User attempts to execute'],
    				then: ['AGENT_NOT_ACTIVE error is returned'],
    			},
    		],
    		examples: [
    			{
    				key: 'basic-execute',
    				input: { agentId: 'agent-123', input: { message: 'Hello' } },
    				output: { runId: 'run-456', status: 'pending', estimatedWaitMs: 5000 },
    			},
    		],
    	},
    });