LangChain Agent Outcome Telemetry

Implementation evidence

LangChain Agent Outcome Telemetry: from boundary to verified row

Use LangChain Agent Outcome Telemetry at a controlled application boundary, keep the event contract small, and verify a known outcome before building aggregate views.

1
Choose the outcome
LangChain agent completion monitoring
2
Define the contract
run_id, agent_name, feature, model, status, and prompt_version
3
Instrument the boundary
Use LangSmith or OpenTelemetry for detailed traces and Telemetry for bounded outcome events that join to product and operational data.
4
Verify the evidence
Exercise a known fixture, then inspect agent_run_completed for one correctly typed terminal row.

Before you start

Prerequisites and boundaries

A server-side TELEMETRY_API_KEY
Stable agent, feature, tool, and prompt-version names
A redaction policy that keeps messages, tool arguments, and outputs out of general events

Delivery setup

Install and initialize server-side

Import telemetry-sh in server-only code and initialize it once with process.env.TELEMETRY_API_KEY. Keep ingestion credentials out of browser bundles, client-visible environment variables, source control, logs, and exception messages.

langchain-install

npm installation

bash

npm install telemetry-sh

1Prepare one reusable server-side delivery client with bounded network behavior.
2Add the outcome event at the success, failure, retry, or timeout boundary.
3Send controlled fixtures and inspect the stored rows before enabling an alert.

Snippet

Start with one structured event

Add this shape where the workflow completes, fails, or retries. Then build the dashboard from real fields.

langchain

LangChain Agent Outcome Telemetry event

javascript

import { randomUUID } from "node:crypto";
import telemetry from "telemetry-sh";

const runId = randomUUID();
const startedAt = performance.now();

try {
  const result = await agent.invoke({
    messages: [{ role: "user", content: userInput }],
  });

  await telemetry.log("agent_run_completed", {
    run_id: runId,
    agent_name: "support_agent",
    feature: "support_resolution",
    prompt_version: "support-v6",
    status: "success",
    duration_ms: Math.round(performance.now() - startedAt),
    tool_calls: countToolCalls(result),
    human_handoff: result.handoffRequired,
    completed_task: result.resolutionCreated,
  });

  return result;
} catch (error) {
  await telemetry.log("agent_run_completed", {
    run_id: runId,
    agent_name: "support_agent",
    feature: "support_resolution",
    prompt_version: "support-v6",
    status: "failed",
    duration_ms: Math.round(performance.now() - startedAt),
    error_type: classifyAgentError(error),
  });
  throw error;
}

Event contract

run_id, agent_name, feature, model, status, and prompt_version

duration_ms, tool_calls, retries, human_handoff, and error_type

accepted, saved, completed_task, or another reviewed product outcome

Implementation checkpoints

Checkpoint 1

Use LangSmith or OpenTelemetry for detailed traces and Telemetry for bounded outcome events that join to product and operational data.

Checkpoint 2

Record a run identifier for correlation, not message content or tool arguments.

Checkpoint 3

Define task success from a terminal product or reviewer outcome rather than the presence of an assistant message.

Verification

Prove the event arrived

Run this after exercising known success and failure cases. Replace the fallback table name if your final event contract differs from the snippet.

langchain-verification

LangChain Agent Outcome Telemetry verification query

sql

SELECT *
FROM agent_run_completed
ORDER BY timestamp_utc DESC
LIMIT 20;

Confirm one terminal row per logical outcome, with the expected status, identifiers, units, and UTC time.

Inspect the inferred schema and verify that retries do not change field types or generate a new logical event ID.

Search the stored fields for credentials, raw payloads, prompts, private content, and unbounded error messages.

Exercise a provider timeout, ingestion rejection, and process shutdown before treating the dashboard as complete.

Implementation references

Review the event contract, data-safety guidance, and upstream primary documentation before enabling a new production path.

Agent task-success SQL Tool-loop SQL Agent observability template LangChain JavaScript agents LangSmith observability

Production boundary

Keep the outcome event small and recoverable

This pattern provides

A bounded, SQL-ready outcome beside the upstream workflow.
Stable fields for dashboards, alerts, and cross-event correlation.
A fixture-driven path for validating success, failure, retry, and timeout behavior.

This pattern does not provide

An OTLP exporter, automatic collection pipeline, or replacement for detailed traces and diagnostic logs.
Exactly-once delivery merely because the payload contains an event ID.
Permission to collect raw provider payloads, user content, credentials, or regulated data.

Delivery and idempotency Batching and shutdown Troubleshoot ingestion

Event schema starting points

Review the row grain, emit boundary, required types, privacy classes, example payload, and validation checklist before adapting a query or snippet to production.

AI agent operations10 fields