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 type { AddressInfo } from 'node:net'
import { RPCServer } from '../server'
describe('RPCServer metrics contract', () => {
it('exposes process and MCP status-heartbeat metrics without runtime auth', async () => {
const server = new RPCServer(0)
await server.start()
const address = (server as unknown as { server: { address(): AddressInfo } }).server.address()
try {
const response = await fetch(`http://127.0.0.1:${address.port}/metrics`)
expect(response.status).toBe(200)
const body = await response.text()
expect(body).toContain('process_cpu_user_seconds_total')
expect(body).toContain('clerum_mcp_status_heartbeat_runs_total')
expect(body).toContain('clerum_mcp_status_heartbeat_in_flight')
expect(body).toContain('clerum_gfs_download_admissions_total')
expect(body).toContain('clerum_gfs_download_transfers_total')
expect(body).toContain('clerum_gfs_download_active')
expect(body).toContain('clerum_gfs_download_quota_total')
expect(body).toContain('clerum_gfs_download_expiry_total')
} finally {
await server.stop()
}
})
})