Documentation for automated readers
A curated documentation index is available at: https://grafana.com/llms.txt
A complete documentation index is available at: https://grafana.com/llms-full.txt
These indexes can help with page discovery before fetching individual documents.
This page is also available in Markdown, which may be easier for automated readers and AI tools to parse than HTML. The Markdown version is available at https://grafana.com/docs/grafana-cloud/observe-and-act/agent-observability/privacy-and-security/agento11y-cli-usage-statistics.md, or by sending Accept: text/markdown to https://grafana.com/docs/grafana-cloud/observe-and-act/agent-observability/privacy-and-security/agento11y-cli-usage-statistics/. For broader documentation discovery, the curated index is available at https://grafana.com/llms.txt and the complete index is available at https://grafana.com/llms-full.txt.
Understand agento11y CLI usage statistics
The agento11y CLI sends anonymous usage statistics about its own commands to Grafana by default. These reports describe agento11y commands, never operations performed by a coding agent. CLI usage reporting is separate from the coding-agent session data sent to your Grafana Cloud stack, including when you use the CLI in local mode.
Before you begin
- Install the
agento11yCLI. Refer to Instrument coding agents.
Understand what is collected
Each report describes one CLI invocation. It contains these fields:
| Fields | What they describe |
|---|---|
service, version, os, arch | The CLI name and version, operating system, and CPU architecture. |
install_id, install_id_persisted | A random installation ID and whether it was saved for future invocations. If it cannot be saved, the CLI uses a temporary ID for that invocation. |
command, surface, agent, skill, flags | The recognized agento11y command, whether it ran as a command, launcher, or help request, a recognized launcher or bundled skill name when applicable, and recognized flag names. |
outcome, exit_code, error_kind, duration_ms | The CLI invocation’s result, exit code, broad error category, and duration. For a launcher, duration covers only startup before it hands control to the coding agent. The CLI doesn’t know the coding agent’s eventual outcome. |
is_tty, is_ci, ci_provider | Whether the CLI ran in a terminal or CI, and a recognized CI provider name when applicable. |
The installation ID is a randomly generated UUID, not derived from your hardware, account, or username. The CLI stores it in its state directory as install-id when that directory is available. You can delete the file to reset the ID.
Reports don’t include user or account IDs, argument or flag values, session content, prompts, tool calls, tool results, endpoint URLs, credentials, or error messages. The agent field names only a recognized agento11y launcher; it doesn’t describe what the coding agent did. Unknown command names aren’t sent verbatim.
Understand how reporting works
The CLI sends at most one JSON report per invocation to https://stats.grafana.org/agento11y-usage-report. Sending is best effort, with a one-second timeout and no retry. Coding-agent hooks, the local server, and agento11y --version don’t send usage reports.
Inspect or turn off reporting
To print the report for an invocation to standard error without sending it, run:
AGENTO11Y_ANONYMOUS_USAGE_STATS=log agento11y doctorTo turn reporting off in your shell, run either command:
export AGENTO11Y_ANONYMOUS_USAGE_STATS=disabledexport DO_NOT_TRACK=1You can also add this line to ~/.config/agento11y/config.env:
AGENTO11Y_ANONYMOUS_USAGE_STATS=disabledThe CLI checks the setting in your shell first, then DO_NOT_TRACK, then config.env. A shell setting of AGENTO11Y_ANONYMOUS_USAGE_STATS=enabled overrides DO_NOT_TRACK=1. These settings control anonymous CLI usage statistics, not the coding-agent session data sent to your Grafana Cloud stack. Refer to Instrument coding agents to configure session content capture.
Next steps
Was this page helpful?
Related resources from Grafana Labs


