Command execution

Command execution runs a specific instruction on the remote host through the key-based SSH connection. When a trigger fires, such as a schedule or an upstream event, a Beam agent selects the command from the approved list, runs it under the granted key, and reads the exit status and output it returns. It writes that outcome to the destination record so the result is visible. Where a command returns an error code, produces output the rule does not recognize, or falls outside the approved list, the agent stops and escalates the run to a person.

Output and log reading

Output and log reading pulls the text a command or file returns from the server so it can be acted on. A Beam agent runs the approved read command, applies the rule that looks for the values or patterns that matter, and records the parsed result against the relevant record. This turns raw server output into a status a workflow can use. When the output does not match any expected pattern, a log is missing, or a reading looks inconsistent with the last run, the agent notes the anomaly and routes it to a person.

Scheduled maintenance runs

Scheduled maintenance runs carry out routine server jobs, such as a service restart or a disk cleanup, on a set cadence. A Beam agent triggers the approved script over SSH, reads the confirmation the server returns, and records whether the job finished cleanly. It keeps a trail of each run against the host record. Where a job reports a failure, exceeds its expected runtime, or would touch something outside the approved scope, the agent halts the sequence and hands the situation to an engineer rather than retrying on its own.