Quickstart: Telemetry

Run the telemetry CLI to see recent usage data and cost savings from your Attune AI setup.

python -m attune.telemetry show

You should see a table of recent telemetry entries logged to help_queries.jsonl.

Prerequisites

• The project is cloned and installed locally
• Redis is accessible (required for heartbeat and coordination features)

Step 1: View recent telemetry entries

from attune.telemetry import UsageTracker

tracker = UsageTracker()
tracker.show()

This prints the most recent entries from the telemetry log.

Step 2: Check cost savings

from attune.telemetry import UsageTracker

tracker = UsageTracker()
tracker.savings()

Expected output: a summary of cost savings from Sonnet → Opus fallback decisions and prompt cache hits.

Step 3: Monitor agent heartbeats

from attune.telemetry import HeartbeatCoordinator

coordinator = HeartbeatCoordinator()
active = coordinator.get_active_agents()

for agent in active:
    print(agent.agent_id, agent.status, agent.progress)

Expected output: one line per active agent showing its ID, current status (running, completed, etc.), and progress as a float between 0.0 and 1.0. An empty list means no agents are currently registered.

Step 4: Request a human approval gate

from attune.telemetry import ApprovalGate

gate = ApprovalGate(agent_id="my-agent")
response = gate.request_approval(
    approval_type="deploy",
    context={"target": "production"},
    timeout=120.0,
)

print(response.approved, response.reason)

Expected output: True <reason> once a human calls gate.respond_to_approval(...), or a timeout response after 120 seconds.

What you just did

• Ran the telemetry CLI to confirm the log is populated
• Queried usage data and cost savings
• Listed active agents via HeartbeatCoordinator
• Opened a human approval gate with ApprovalGate

Next: read the telemetry concept page — say "what is Attune telemetry?" — to understand how UsageTracker, HeartbeatCoordinator, CoordinationSignals, and ApprovalGate fit together in a multi-agent workflow.