Logs & activity

The console's log viewer, activity feed, and credit usage for an agent.

Open an agent to find three views into what it's doing: raw logs, a higher-level activity feed of events, and credit usage.

This built-in visibility is part of what running an agent gets you over building it yourself — you don't need to add your own logging to see what the agent has been doing.

Logs

The Logs tab shows the most recent output from the agent — useful for debugging channel connections and errors.

Activity

The Activity tab shows a feed of recent events: messages handled, channel connects, and errors.

Credit usage

The Credits tab reports the credit usage attributed to this agent, so you can see which agent is spending your allowance.

What you're actually looking for

In practice these three tabs answer two very different questions.

Debugging a bad response

A customer says the WhatsApp bot gave a wrong answer. Start in Activity to find the message and confirm which channel and roughly when it happened, then switch to Logs around that timestamp for the raw output — this is where you'll spot a failed channel reconnect, a model error, or a request that never reached the provider at all. If the issue traces back to how the agent is set up rather than a one-off blip, the fix usually lands on the Configuration screen — a stale credential, the wrong model selected, or memory being off when the customer expected it remembered.

Watching credit usage climb

If an agent's usage looks higher than expected — a Starter plan burning through its 25 credits faster than a normal month — the Credits tab shows whether it's trending up steadily or spiking on a particular day, and Activity shows the message volume behind it. That combination tells you whether it's just organic growth in conversations (a case for upgrading the plan) or a loop of repeated messages worth investigating in Logs.