Skip to content
CyberSimStart training

Investigation playbook · endpoint

Endpoint detection playbook

For detections on a computer: suspicious processes, scripts, persistence, credential access.

Answer these before you decide

  1. What ran, and what started it? Read the parent-child chain (Office or a browser spawning a shell is a red flag).
  2. Decode what you can: base64 command lines, download cradles, LOLBins such as rundll32, regsvr32, certutil.
  3. Was it blocked or did it run? "Quarantined" and "Executed" are very different nights.
  4. Is it expected: an admin tool by an admin from the jump host, a test file, a software rollout in the handover?
  5. Did it talk to the network or touch credentials (LSASS access, new services, scheduled tasks)?

When it is an incident: first containment

Isolate the host when something is executing; escalate with the process tree.

A playbook says what to check. Whether this alert is real is what the logs tell you.

Events to read

ATT&CK techniques these alerts carry

Practise it

Alerts in the shift that use this playbook

In the glossary

Other playbooks