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 { isTraceContextV1, mintChannelTraceContext } from '../traceContext'
import type { Message } from '../types'
const message: Message = {
channelType: 'telegram',
channelId: '424242',
sender: '123456',
content: 'hello',
timestamp: new Date('2026-07-11T08:00:00.000Z'),
messageId: '9001',
threadId: 'thread-1',
providerIdentity: {
medium: 'telegram',
providerUserId: '123456',
providerChannelId: '424242',
providerEventId: 'telegram:424242:9001',
},
}
describe('TraceContextV1 channel-reader contract', () => {
it('mints the mcp-host-compatible channel context', () => {
const context = mintChannelTraceContext(message)
expect(context).toMatchObject({
version: 1,
origin: 'channel_event',
sessionId: null,
correlationRefs: [
'provider-event:telegram:424242:9001',
'channel-delivery:9001',
'provider-thread:thread-1',
],
})
expect(context.runId).toMatch(
/^[0-9a-f]{8}-[0-9a-f]{4}-4[0-9a-f]{3}-[89ab][0-9a-f]{3}-[0-9a-f]{12}$/i
)
expect(isTraceContextV1(context)).toBe(true)
})
it('rejects values outside the local wire contract', () => {
expect(isTraceContextV1({ ...mintChannelTraceContext(message), unexpected: 'value' })).toBe(
false
)
})
})