Evenfire is an intelligence layer platform for running LLM agents that take real actions, on your infrastructure
TypeScript
import { describe, expect, it } from 'vitest'
import { buildPromptBridgeUsageEvent } from './index'
const WORKFLOW_RUN_ID = '00000000-0000-4000-8000-000000000001'
describe('PluginWorkloadSdk server usage attribution', () => {
it('does not emit invalid workflow usage when promptBridge lacks authoritative run metadata', () => {
const event = buildPromptBridgeUsageEvent({
binding: {
hostRef: 'sandbox-recipes/risk-review',
recipeNamespace: 'sandbox-recipes',
recipeName: 'risk-review',
},
provider: 'openai',
model: 'gpt-4o',
inputTokens: 10,
outputTokens: 5,
callerRef: 'worker',
})
expect(event).toBeNull()
})
it('emits the accepted workflow usage shape when promptBridge carries run and secret metadata', () => {
const event = buildPromptBridgeUsageEvent({
binding: {
hostRef: 'sandbox-recipes/risk-review',
recipeNamespace: 'sandbox-recipes',
recipeName: 'risk-review',
},
provider: 'openai',
model: 'gpt-4o',
inputTokens: 10,
outputTokens: 5,
callerRef: 'worker',
metadata: {
workflowExecutionId: `${WORKFLOW_RUN_ID}:risk-review:2026-07-12T12:00:00.000Z`,
workflowRunId: WORKFLOW_RUN_ID,
llmSecretName: 'chatllm-api-keys',
workflowTeamId: '11111111-1111-4111-8111-111111111111',
workflowUserId: '22222222-2222-4222-8222-222222222222',
},
})
expect(event).toMatchObject({
run_id: WORKFLOW_RUN_ID,
host_ref: 'sandbox-recipes/risk-review',
recipe_name: 'risk-review',
source_kind: 'workflow',
llm_secret_name: 'chatllm-api-keys',
task_id: `${WORKFLOW_RUN_ID}:risk-review:2026-07-12T12:00:00.000Z`,
team_id: '11111111-1111-4111-8111-111111111111',
user_id: '22222222-2222-4222-8222-222222222222',
sender: 'worker',
channel_type: 'plugin_workload_sdk',
input_tokens: 10,
output_tokens: 5,
})
})
it('emits a server-bound SDK-only usage shape without inventing a workflow run', () => {
const invocationId = '00000000-0000-4000-8000-000000000099'
const event = buildPromptBridgeUsageEvent({
binding: {
hostRef: 'sandbox-recipes/stepless',
recipeNamespace: 'sandbox-recipes',
recipeName: 'stepless',
},
runtimeMode: 'sdk-only',
invocationId,
provider: 'openai',
model: 'gpt-5.4-mini',
inputTokens: 10,
outputTokens: 5,
callerRef: 'backend-worker',
promptBridgeMetadata: {
targetRef: 'openai-primary',
credentialSlot: 'openai-api-key',
fallbackUsed: false,
attemptCount: 1,
},
metadata: { llmSecretName: 'openai-api-key' },
})
expect(event).toMatchObject({
source_kind: 'plugin_workload_sdk',
channel_type: 'plugin_workload_sdk',
recipe_name: 'stepless',
run_id: null,
task_id: null,
llm_secret_name: 'openai-api-key',
prompt_bridge_metadata: { invocation_id: invocationId },
})
})
it('records actual promptBridge target attribution without credential values', () => {
const event = buildPromptBridgeUsageEvent({
binding: {
hostRef: 'plugin-workload-sdk/sandbox-app',
recipeNamespace: 'sandbox-apps',
recipeName: 'prompt-notify',
},
runtimeMode: 'sdk-only',
invocationId: '00000000-0000-4000-8000-000000000100',
provider: 'openai',
model: 'gpt-5.4-mini',
inputTokens: 10,
outputTokens: 5,
callerRef: 'backend-worker',
promptBridgeMetadata: {
targetRef: 'fallback-openai',
credentialSlot: 'openai-api-key',
fallbackUsed: true,
attemptCount: 2,
},
})
expect(event).toMatchObject({
provider: 'openai',
model: 'gpt-5.4-mini',
prompt_bridge_metadata: {
target_ref: 'fallback-openai',
credential_slot: 'openai-api-key',
fallback_used: true,
attempt_count: 2,
},
})
expect(JSON.stringify(event)).not.toContain('credentialTicket')
expect(JSON.stringify(event)).not.toContain('secret-')
})
it('does not emit reporter usage for oauth-broker subscriptions', () => {
expect(
buildPromptBridgeUsageEvent({
binding: {
hostRef: 'sandbox-recipes/risk-review',
recipeNamespace: 'sandbox-recipes',
recipeName: 'risk-review',
},
provider: 'codex-subscription',
model: 'gpt-5.1',
inputTokens: 10,
outputTokens: 5,
callerRef: 'worker',
metadata: {
workflowExecutionId: `${WORKFLOW_RUN_ID}:risk-review:2026-07-12T12:00:00.000Z`,
workflowRunId: WORKFLOW_RUN_ID,
llmSecretName: 'chatllm-api-keys',
},
})
).toBeNull()
expect(
buildPromptBridgeUsageEvent({
binding: {
hostRef: 'plugin-workload-sdk/sandbox-app',
recipeNamespace: 'sandbox-apps',
recipeName: 'prompt-notify',
},
runtimeMode: 'sdk-only',
invocationId: '00000000-0000-4000-8000-000000000100',
provider: 'codex-subscription',
model: 'gpt-5.1',
inputTokens: 10,
outputTokens: 5,
callerRef: 'backend-worker',
promptBridgeMetadata: {
targetRef: 'codex-primary',
credentialSlot: 'codex-oauth',
fallbackUsed: false,
attemptCount: 1,
},
})
).toBeNull()
expect(
buildPromptBridgeUsageEvent({
binding: {
hostRef: 'plugin-workload-sdk/sandbox-app',
recipeNamespace: 'sandbox-apps',
recipeName: 'prompt-notify',
},
runtimeMode: 'sdk-only',
invocationId: '00000000-0000-4000-8000-000000000100',
provider: 'grok-subscription',
model: 'grok-4.6',
inputTokens: 10,
outputTokens: 5,
callerRef: 'backend-worker',
})
).toBeNull()
})
})