GitHub Copilot CLI Integration

GitHub Copilot CLI

GitHub Copilot CLI is GitHub's agentic coding tool that runs in your terminal. With the Dash0 agent plugin it emits OpenTelemetry traces about every session — token usage, cost, the model behind each turn, tool calls, and the pull requests it helps ship. Sending that telemetry to Dash0 gives engineering leaders and platform teams a single place to understand how Copilot CLI is adopted, what it costs, and the impact it has — without changing how developers work.

Use Cases

Cost visibility: Track spend by user, team, and model so you can understand and control what Copilot CLI costs.
Adoption tracking: Follow active users, new adopters, and usage distribution to see how Copilot CLI rolls out across your engineering teams.
Productivity signals: Correlate sessions, tool calls, and assisted pull requests to understand the impact Copilot CLI has on your team's output.

Overview

GitHub Copilot CLI ships native OpenTelemetry that carries the token, cost, model, and tool-execution detail of each session. The Dash0 agent plugin turns that into canonical per-turn spans and exports them to Dash0, where the AI Coding section reports what Copilot CLI costs and how widely your teams use it.

Prerequisites

GitHub Copilot CLI installed and authenticated on macOS, Linux, or Windows, on amd64 or arm64.
A Dash0 auth token with ingest permissions.
Your name and email set in your git configuration. The plugin uses them to attribute activity to users:

Install the plugin

Add the Dash0 marketplace, install the plugin from it, and restart copilot so the hooks load at startup:

Configure

Run the configure command inside Copilot CLI:

When prompted, provide your Dash0 connection details:

OTLP URL: {{endpoint_otlp_http}}
Auth token: {{token}}
Dataset (optional): {{dataset}}

This writes your configuration to ~/.copilot/dash0-agent-plugin.local.md (%USERPROFILE%\.copilot\dash0-agent-plugin.local.md on Windows), or to .copilot/dash0-agent-plugin.local.md in the current workspace if you choose project scope. It also installs a launch wrapper that enables Copilot's native OpenTelemetry into a per-session file, which is where the per-turn token, cost, and model usage comes from. The skill asks which shell you launch copilot from and installs the matching wrapper:

bash or zsh: a copilot function in ~/.bashrc or ~/.zshrc.
PowerShell: a copilot function in the file $PROFILE names.
cmd.exe: a copilot.cmd wrapper in a PATH directory that comes before the npm global bin (%APPDATA%\npm).

Open a new shell afterward so the wrapper takes effect.

For more information on choosing a dataset, see the dataset documentation.

Attribute Usage by Team (Optional)

Copilot CLI attaches the authenticated user to every span, so cost and adoption break down by user and model out of the box. To also break down by team or department, provide a team name when you run /dash0-configure. It goes into the plugin config and is attached to every span.

Verify

Open a new shell and run copilot, then issue any command.
Within a few seconds, the telemetry appears in the AI Coding section of Dash0.

A copilot launched from a shell without the launch wrapper still emits spans, but without the per-turn usage detail.

On Windows, check that the wrapper is in place with Get-Command copilot -CommandType Function in PowerShell, or where copilot.cmd in cmd.exe. A PowerShell profile that fails to parse loads nothing at all, which leaves the function undefined and the spans without usage data.

Technology

GitHub Copilot CLI

Overview