An Agent is an AI process that Claudius Corp spawns to execute a workflow phase. Each agent runs independently and communicates back to Claudius Corp when it completes its work.
Viewing Agents #
Click Agents in the sidebar to see all agents — running, completed, and historical.
For each agent you can see:
– Which project and phase it’s working on
– Current status (Starting, Running, Completed, Crashed, etc.)
– Start and end times
– API cost (tokens used)
Agent Statuses #
| Status | Meaning |
|---|---|
| Starting | Agent process is being launched |
| Running | Agent is actively working |
| Waiting | Agent is waiting for your input |
| Completed | Agent finished successfully |
| Crashed | Agent exited with an error |
| Terminated | You stopped the agent manually |
| Paused | Agent is paused with the project |
Viewing Agent Logs #
Click any agent to open its detail page. The log tab shows everything the agent printed to its output — useful for understanding what it did or diagnosing problems.
Stopping an Agent #
From the agent detail page, click Stop Agent. The agent will be terminated. The phase will be marked as failed and you can retry it.
Note: Stopping an agent does not undo any changes the agent has already made (e.g. files written or commits made).