GitHub Copilot · OpenTelemetry

GitHub Copilot OpenTelemetry

A practical guide to GitHub Copilot telemetry: available metrics, events and traces, configuration, privacy considerations, and what engineering teams can learn from the data.

Updated September 2026 · 12 min read

GitHub Copilot

Coding agents

OpenTelemetry

Export layer

Signals

Metrics
Events
Traces

Compatible backend

Your observability stack

Quick answer

What is GitHub Copilot OpenTelemetry?

GitHub Copilot supports OpenTelemetry for exporting observability data from Copilot coding agents. This telemetry can provide visibility into agent activity, model interactions, tool usage, token consumption and execution behavior.

OpenTelemetry-compatible backends can collect this data so engineering organizations can analyze how AI coding agents operate across development workflows.

In short: OpenTelemetry makes supported GitHub Copilot agent activity observable outside Copilot itself.

Telemetry signals

What Telemetry Does GitHub Copilot Provide?

GitHub documents three types of data that Copilot clients can send to an OpenTelemetry-compatible backend.

Available
Metrics
Quantitative telemetry including AI model token consumption and supported agent activity, such as tool call counts and agent durations.
Available
Events
Structured OpenTelemetry events describing supported Copilot agent activity, recorded at a specific point in time.
Available
Traces
Distributed traces showing agent operations, including supported model and tool calls within an agent session.

Telemetry reference

GitHub Copilot OpenTelemetry Data

GitHub Copilot telemetry exposes signals that can help organizations understand how coding agents interact with models and tools during development workflows. Signal names below follow the OpenTelemetry GenAI semantic conventions and the Copilot-specific namespaces documented by GitHub and the Copilot clients.

GitHub Copilot OpenTelemetry signals, official identifiers and what they help observe
Telemetry	Signal	Official identifier	What it helps observe
Agent activity	Span	`invoke_agent`	How supported Copilot agents execute development tasksThe root span of an agent session wraps the whole orchestration, including every model call and tool execution.
Model calls	Span	`chat`	Interactions between the agent and AI modelsOne span per model API call, carrying the requested and resolved model, finish reasons and latency.
Token consumption	Histogram metric	`gen_ai.client.token.usage`	AI model token usageInput and output token counts follow the OpenTelemetry GenAI semantic conventions.
Model call duration	Histogram metric	`gen_ai.client.operation.duration`	Model API latencyDuration of model API calls, in seconds.
Tool calls	Span	`execute_tool`	Tools invoked by the agent during executionOne span per tool invocation, with the tool name, tool type (function or MCP extension) and call identifier.
Tool call volume	Counter metric	`copilot_chat.tool.call.count`	Tool invocations by name and successCounts tool invocations so teams can see which tools agents rely on.
Session activity	Counter metric	`copilot_chat.session.count`	Chat sessions startedCounts chat sessions as they begin.
Model call details	Event	`gen_ai.client.inference.operation.details`	Full model call metadataRecords model, tokens and finish reason for a model call.

Common attributes

gen_ai.operation.name
Operation type, such as invoke_agent, chat or execute_tool.
gen_ai.agent.name
Agent name, for example copilot, copilotcli or claude.
gen_ai.request.model
Model requested for the call.
gen_ai.usage.input_tokens
Input tokens consumed.
gen_ai.usage.output_tokens
Output tokens produced.
gen_ai.tool.name
Name of the invoked tool, for example readFile.
github.copilot.agent.type
Whether the agent is builtin, custom or plugin.
error.type
Error class, emitted on failure.

Use cases

What Can Engineering Teams Learn From GitHub Copilot Telemetry?

Copilot telemetry can help engineering organizations answer questions about agent usage, AI consumption, tooling and execution behavior.

Agent Usage
Understand how supported Copilot coding agents are being used across engineering workflows.
AI Consumption
Analyze model interactions and token consumption to understand how AI resources are used.
Tools & Workflows
Observe which tools agents invoke and how those tools participate in development tasks.
Agent Performance
Use traces and telemetry to investigate execution behavior, latency and errors across agent workflows.
Activity and token metrics provide context about AI usage. They should not be treated as standalone measures of developer productivity.

How it works

How GitHub Copilot OpenTelemetry Works

Supported GitHub Copilot agents can generate OpenTelemetry data that is exported to compatible observability infrastructure.

GitHub Copilot

OpenTelemetry

Metrics
Events
Traces

OTLP

HTTP or gRPC

Collector or compatible backend

Read the official GitHub Copilot OpenTelemetry documentation

Configuration

How to Configure OpenTelemetry for GitHub Copilot

GitHub provides configuration options for exporting supported Copilot telemetry to OpenTelemetry-compatible infrastructure. Telemetry is off by default and activates once it is enabled or an OTLP endpoint is configured.

# Enable OpenTelemetry export from the Copilot client
export COPILOT_OTEL_ENABLED=true

# Send OTLP data to your collector or compatible backend
export OTEL_EXPORTER_OTLP_ENDPOINT="http://localhost:4318"
export OTEL_EXPORTER_OTLP_PROTOCOL="http/protobuf"

# Authenticate against the backend, when required
export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Bearer <token>"

View advanced configuration

Documented client settings and environment variables. Environment variables take precedence over client settings, and enterprise-managed Copilot settings take precedence over both.

github.copilot.chat.otel.enabled — VS Code setting that enables OTel emission. Default false.
github.copilot.chat.otel.exporterType — Exporter to use: otlp-http, otlp-grpc, console or file. Default "otlp-http".
github.copilot.chat.otel.otlpEndpoint — OTLP collector endpoint. Default "http://localhost:4318".
COPILOT_OTEL_ENDPOINT — OTLP endpoint URL that takes precedence over OTEL_EXPORTER_OTLP_ENDPOINT.
COPILOT_OTEL_PROTOCOL — Overrides the OTLP protocol (grpc or http).
COPILOT_OTEL_CAPTURE_CONTENT — Captures full prompt and response content. Default false.
COPILOT_OTEL_MAX_ATTRIBUTE_SIZE_CHARS — Maximum characters per content attribute. 0 disables truncation.
OTEL_SERVICE_NAME — Service name in resource attributes. Default "copilot-chat".
OTEL_RESOURCE_ATTRIBUTES — Extra resource attributes, for example team.id=platform,department=engineering.

For enterprise rollout, backend examples and client-specific behavior, refer to the client documentation.

Privacy

Privacy and Sensitive Data

Coding-agent telemetry can reveal detailed information about development workflows. Organizations should understand what data is exported before enabling telemetry at scale.

Prompts & Content
GitHub documents that, by default, the data does not include prompts, responses or tool arguments. Content capture is opt-in through the captureContent setting or COPILOT_OTEL_CAPTURE_CONTENT, and GitHub warns it may contain sensitive information such as code, file contents and user prompts.
Tool Activity
Tool telemetry records the tool name, tool type and call identifiers, plus attributes such as edit type or a hashed MCP server name. File paths, shell commands and tool arguments or results are only emitted when content capture is enabled.
Telemetry Destination
Telemetry is exported over OTLP to an endpoint your organization chooses, either directly or through an OpenTelemetry Collector. Enterprises can enforce that configuration centrally with Copilot managed settings.

Good practice: Collect only the telemetry required to answer your organization's observability questions.

From telemetry to analytics

OpenTelemetry Gives You the Data. What Do You Do With It?

GitHub Copilot OpenTelemetry provides detailed technical signals about agent execution, but raw telemetry does not automatically explain how AI-assisted development is evolving across an engineering organization.

Engineering teams still need to structure these signals to understand broader patterns around adoption, AI consumption, tools and development workflows.

Turn GitHub Copilot Telemetry Into Engineering Intelligence

helloMetry transforms supported AI coding telemetry into structured analytics designed to help engineering organizations understand how AI agents are being adopted and used.

Adoption
AI Consumption
Agent Activity
Workflows

Explore GitHub Copilot with helloMetry

Overview

Last 7 days

2.842

AI Sessions+18%

312

Skill Usage+12%

18.672

Invocations+23%

€430

Est. Cost-8%

Activity over time

09:0013:1517:30

Time saved

156 h+26%

Tasks automated

1.284+19%

Skills breakdown

Code38%
Analysis24%
Writing18%
Research12%
Other8%

Active skills

48+9%

Success rate

96.2%+2%

Top skills

/code_review142 usages4.892
/debug98 usages3.210
/refactor85 usages2.761
/data_analysis74 usages2.104
/summarize68 usages1.874

Explore Other OpenTelemetry Integrations

Claude Code
Claude Code OpenTelemetry: metrics, events and traces, configuration and privacy considerations.
Read the guide
Gemini CLI
Gemini CLI OpenTelemetry: available metrics, logs and traces, setup and privacy considerations.
Read the guide
More integrations
Coming soon — additional OpenTelemetry guides for other AI coding agents.

FAQ

Frequently Asked Questions

Does GitHub Copilot support OpenTelemetry?

Yes. GitHub provides OpenTelemetry support for supported Copilot coding-agent telemetry, allowing observability data to be exported to compatible infrastructure.

What GitHub Copilot data can OpenTelemetry expose?

Depending on the supported agent and configuration, telemetry can provide visibility into agent operations, model calls, tool activity, token consumption and execution traces.

Can GitHub Copilot token usage be monitored?

GitHub Copilot OpenTelemetry includes supported metrics related to AI model token consumption.

Does GitHub Copilot provide OpenTelemetry traces?

Yes. Supported Copilot agent telemetry includes distributed traces that can represent agent operations and interactions with models and tools.

Does helloMetry replace OpenTelemetry?

No. OpenTelemetry provides the standard used to collect and export telemetry. helloMetry uses supported telemetry to transform AI coding-agent data into engineering analytics.

Is helloMetry an employee monitoring tool?

helloMetry is designed to help organizations understand AI adoption, usage, costs and engineering workflows rather than rank developers using simplistic activity metrics.

Sources & further reading

Primary technical source: official GitHub documentation

GitHub Docs — OpenTelemetry for GitHub Copilot

Client documentation

VS Code — Monitor agent usage with OpenTelemetry

GitHub Copilot OpenTelemetry

Metrics

Events

Traces

Common attributes

Agent Usage

AI Consumption

Tools & Workflows

Agent Performance

Prompts & Content

Tool Activity

Telemetry Destination