Read a sample: 1. What an Incident Report Must Do
An investigation may begin with a suspicious alert, but it rarely ends with one neat answer. Analysts review records, ask questions, test explanations, and discover gaps. Their working notes capture that process as it unfolds. An incident report has a different job: it turns the useful results of that work into an account another person can understand, evaluate, and act on.
That distinction matters. A collection of accurate notes can still leave a reader unsure what happened. A polished summary can sound convincing while concealing weak evidence. A useful report sits between those failures. It is selective without being misleading, direct without overstating certainty, and detailed enough for its intended readers to follow the reasoning.
This book focuses on that translation. You do not need access to a commercial security platform or a live network to practice it. The essential materials are observations, sources, questions, and decisions about how to communicate them.
An incident report records what the investigation found at a particular point in time. It explains the relevant sequence, identifies the evidence behind important claims, describes known effects, and makes clear what remains unresolved. When appropriate, it also records response actions and recommends further work.
Different readers use that account for different purposes. A security analyst may need to understand why an account was considered compromised and which records support that conclusion. An IT manager may need to assign follow-up work. A business leader may need to understand whether operations were disrupted and what decisions are outstanding. A reviewer who was not present may need to reconstruct why the team acted as it did.
A report does not need to answer every imaginable question. It does need to make its boundaries visible. If the team examined one account and one application, the report should not imply that the entire environment was reviewed. If a file was accessed but its contents and destination are unknown, the report should not quietly convert access into confirmed data theft—or into proof that nothing was taken.
A practical incident report helps readers answer several questions:
- What happened, or what is believed to have happened?
- How was the activity discovered, and what period was examined?
- Which accounts, systems, or information were within the investigation’s scope?
- What evidence supports the important findings, and where are the gaps?
- What impact and response actions are confirmed?
- What decisions, validation, or further work remain?
Not every report needs a separate heading for each question. The point is that readers should not have to guess at the answers that matter to their role.
Consider the following fictional incident. It will recur throughout this book as a shared set of facts to organize, test, and communicate. The company, people, systems, times, and records are hypothetical.
On a Tuesday morning, the security team at Harborview Design receives an alert about an unusual sign-in to a cloud file-sharing application. The alert concerns the account of Maya Chen, a project coordinator. The sign-in record shows that the account authenticated at 08:42 UTC from a location not normally associated with Maya’s recent activity. At 08:47 UTC, the application recorded access to a project folder. The available activity record does not establish whether any files were downloaded.
At 09:10 UTC, an analyst contacts Maya. She says she did not initiate the sign-in and was working from her usual location that morning. At 09:25 UTC, an administrator disables the account’s active sessions and resets its credentials. The team later confirms that Maya can sign in again using the organization’s approved recovery process.
The team reviews the available sign-in and application activity for Maya’s account during the period surrounding the alert. It has not yet established how the sign-in occurred. It also has not reviewed every account or application in the organization. The project-folder owner has been asked to identify the information in the folder and assess possible consequences if it was viewed. That assessment is still pending.
Even this short scenario contains several distinct kinds of information. The sign-in and folder-access entries are recorded observations. Maya’s statement is a relevant account from a person, not a system log. The analyst’s concern that someone else used the account is an inference supported by the unusual activity and Maya’s statement. The method of access, whether files were downloaded, and the consequences of possible viewing remain unknown.
A report should preserve those differences. It should also avoid treating the alert itself as proof of compromise. An alert identifies activity worth examining; it does not necessarily establish what the activity means.
Working notes often reflect the order in which an analyst learned things, not the order in which events occurred. They may contain shorthand, repeated observations, possible explanations, corrections, and reminders. This is normal. Notes are a workspace, not a finished argument.
A simplified set of notes for the Harborview case might look like this:
> Alert on Maya account—unusual location. Sign-in 08:42 UTC. Folder activity 08:47. Maya says not her; called at 09:10. Admin ended sessions/reset credentials at 09:25. Download? App record unclear. Ask folder owner what was in it. How did they get in? Check sign-in records. Maya recovered access.
These notes are useful to the analyst, but several questions remain for a reader. Which items came from application records? Was the folder activity confirmed or merely suspected? Did the administrator act before or after Maya was contacted? Does “recovered access” mean…