Anti-Forensics Field Manual

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== Registry Hives ======

===== Definition =====
Registry Hives is a Windows registry-related artifact or artifact family used to interpret configuration, user activity, device history, persistence, or application state.
In a forensic report, registry hives should be treated as a source of evidence and uncertainty, not as a shortcut to intent. The useful question is what the record supports, what it does not support, and what another source says when asked the same question.

===== Background =====
The registry is a useful witness with a bad filing system. It remembers things unevenly, timestamps entire rooms instead of individual conversations, and still expects the examiner to be grateful.
For practitioners, the value is not drama; it is knowing which facts can survive review and which ones need to stay in the notes. The registry remembers a lot, but it does not always remember which sentence the examiner wanted.

===== Technical Description =====
Review focuses on hive source, user SID, key path, value names, last-write time, transaction logs, control sets, and corroborating artifacts.
Version, platform, retention, and collection scope should be documented before the artifact is promoted to a finding.

===== Forensic Relevance =====
  * **User activity review:** It may support account or profile context, but only when independent artifacts point the same direction.
  * **Contradiction testing:** It is useful for finding places where logs, metadata, storage behavior, or accounts disagree.
  * **Reporting decisions:** It helps decide how narrow the finding should be when registry hives looks interesting but not conclusive.
  * **Timeline reconstruction:** Registry Hives can help place activity in sequence when the time source and collection conditions are understood.
  * **Recovery analysis:** It can explain why data was recovered, missed, corrupted, or only partially reconstructed.

===== Evidence Sources =====
^ Evidence source ^ What it may show ^ Reliability limits ^ What it cannot prove alone ^
| Registry hives | May show configuration, persistence, service paths, or user activity. | Last-write times may cover multiple values. | Do not prove which value changed alone. |
| Event logs | May corroborate service creation, task execution, or administrative activity. | Coverage depends on policy. | Do not prove intent. |
| File-system metadata | May support binary path, timestamps, and signer context. | Can be rewritten by updates or restores. | Does not prove operator identity. |
| Autorun locations | May show persistence opportunities. | Legitimate software uses them constantly. | Do not make persistence malicious by default. |

===== Interpretation Limits =====
The most common error is treating registry hives as proof of motive. It may support a technical event, a sequence, or a contradiction, but motive needs stronger ground.
Another bad leap is treating absence as intent. Missing data can come from retention, configuration, collection scope, sync behavior, overwriting, media behavior, or ordinary use.
A tool result should be described as a parsed or recovered record, not as the tool's opinion about what happened. Tools surface evidence; they do not understand it.
The registry remembers a lot, but it does not always remember which sentence the examiner wanted.

===== Common Misinterpretations =====
  * Ignoring control sets, transaction logs, and profile context.
  * Treating presence in an autorun location as malicious without software context.
  * Attributing shared-profile registry activity to a person without support.
  * Treating registry hives as proof that a specific user acted, when the artifact only supports system or account context.
  * Treating a timestamp as exact truth without checking timezone, clock drift, and source semantics.

===== Example Scenario =====
A reviewer sees registry hives in the case file and asks whether it actually supports the written conclusion.
The artifact may explain part of the sequence, but ordinary system behavior still needs to be ruled in or out.
The careful next step is to normalize the time source, compare independent artifacts, and write the finding as support rather than proof.

===== Analysis Workflow =====
  - Define the question before opening another parser: what should Registry Hives help answer?
  - Preserve the source evidence and document how the registry hives were collected.
  - Record tool versions, input paths, output paths, time settings, and errors.
  - Identify observed facts before writing any interpretation.
  - Normalize time sources and document timezone, clock drift, and collection-time effects.
  - Compare at least two independent artifact families before raising confidence.
  - Consider benign explanations, automated behavior, retention, sync, and storage-device behavior.
  - Write conclusions proportionally: observed fact first, inference second, uncertainty always visible.

===== Reporting Guidance =====
Reporting on registry hives should be precise enough that another analyst can retrace the claim without inheriting the original examiner's confidence.
Avoid wording that converts possibility into intent. The report should not say a user deliberately deleted, hid, wiped, or tampered with evidence unless the evidence actually supports that conclusion.
Report-ready wording:
  * The available registry hives are consistent with activity related to registry hives, but they do not independently establish motive or user intent.
  * Recovery was limited under the examined conditions; additional corroboration would be required before concluding deliberate destruction.
  * The finding should be read with the collection scope, time-source limits, and alternate explanations described in this report.

===== Confidence and Reliability =====
^ Confidence level ^ What it looks like for this topic ^ How to report it ^
| Low | Registry hives exist, but collection scope, time source, or surrounding context is limited. | State the observation and keep interpretation narrow. |
| Moderate | Registry hives align with Event logs, but attribution or intent remains unresolved. | Say the artifacts support the finding, not that they prove it. |
| High | Multiple independent sources agree on sequence, source system, account context, and collection conditions. | Use stronger language, but still separate observed facts from inference. |

===== Related Concepts =====
  * [[wiki:digital_forensics:windows:windows_registry|Windows Registry]] - Useful when checking whether another artifact supports the same interpretation.
  * [[wiki:digital_forensics:windows:muicache|MUICache]] - Often appears nearby in reviews of the same system or behavior.
  * [[wiki:digital_forensics:windows:ntuser_dat|NTUSER.DAT]] - Often appears nearby in reviews of the same system or behavior.
  * [[wiki:digital_forensics:windows:runmru|RunMRU]] - Helps keep the finding grounded in a wider evidence pattern.
  * [[wiki:digital_forensics:windows:sam_hive|SAM Hive]] - Often appears nearby in reviews of the same system or behavior.
  * [[wiki:digital_forensics:windows:security_hive|SECURITY Hive]] - Compare this concept when testing sequence and corroboration.

===== Tools =====
  * [[https://ericzimmerman.github.io/#!index.md|Eric Zimmerman's Tools]] - Windows artifact parsers including MFTECmd, PECmd, EvtxECmd, and others.
  * [[https://github.com/keydet89/RegRipper3.0|RegRipper]] - Registry parsing framework and plugins.
  * [[https://www.sans.org/tools/kape|KAPE]] - Targeted artifact collection and processing.
  * [[https://learn.microsoft.com/en-us/sysinternals/|Sysinternals Suite]] - Microsoft Windows internals and triage tools.
  * [[https://learn.microsoft.com/en-us/sysinternals/downloads/autoruns|Autoruns]] - Persistence and autorun location review.
  * [[https://learn.microsoft.com/en-us/sysinternals/downloads/procmon|Process Monitor]] - Process, registry, and file activity observation.
  * [[https://learn.microsoft.com/en-us/sysinternals/downloads/sysmon|Sysmon]] - High-fidelity Windows endpoint telemetry.
  * [[https://github.com/Yamato-Security/hayabusa|Hayabusa]] - Windows event log hunting and timeline support.
  * [[https://github.com/WithSecureLabs/chainsaw|Chainsaw]] - Windows event log search and detection.
  * [[https://github.com/wagga40/Zircolite|Zircolite]] - Sigma-based event log scanning.
  * [[wiki:tools:start|Forensic Tools]] - Full tool directory for the wiki.

===== Limitations of Tools =====
Tools parse, surface, and organize evidence. They do not create conclusions.
Parser output can be affected by version differences, unsupported formats, corrupted records, timezone handling, partial collection, and storage behavior outside the tool's view.
When a tool produces a strong-looking result, validate it against another tool or source where the stakes justify it. A parser can recover a fragment; it cannot tell you whether the fragment deserves a paragraph in the report.

===== References and Further Reading =====
  * [[https://learn.microsoft.com/en-us/sysinternals/|Microsoft Sysinternals]] - Windows internals and triage tools.
  * [[https://learn.microsoft.com/en-us/windows/win32/eventlog/event-logging|Microsoft Windows Event Log]] - Event logging reference.
  * [[https://www.forensicfocus.com/articles/windows-forensics-and-security/|Forensic Focus: Windows Forensics and Security]] - Windows artifact overview.
  * [[https://csrc.nist.gov/publications/detail/sp/800-86/final|NIST SP 800-86]] - Guide to integrating forensic techniques into incident response.
  * [[https://www.nist.gov/itl/ssd/software-quality-group/computer-forensics-tool-testing-program-cftt|NIST CFTT]] - Computer Forensics Tool Testing program.

===== See Also =====
  * [[wiki:digital_forensics:windows:start|Windows]]
  * [[wiki:digital_forensics:windows:windows_registry|Windows Registry]]
  * [[wiki:digital_forensics:windows:muicache|MUICache]]
  * [[wiki:digital_forensics:windows:ntuser_dat|NTUSER.DAT]]
  * [[wiki:digital_forensics:windows:runmru|RunMRU]]
  * [[wiki:digital_forensics:windows:sam_hive|SAM Hive]]
  * [[wiki:digital_forensics:windows:security_hive|SECURITY Hive]]
  * [[wiki:digital_forensics:windows:shellbags|ShellBags]]
  * [[wiki:digital_forensics:windows:software_hive|SOFTWARE Hive]]
  * [[wiki:digital_forensics:windows:system_hive|SYSTEM Hive]]
  * [[wiki:digital_forensics:windows:typedpaths|TypedPaths]]
  * [[wiki:tools:start|Forensic Tools]]
  * [[wiki:resources:start|Project Resources]]

===== Reader Takeaway =====
Registry Hives is useful when it helps explain what the evidence can support and where the limits begin. Treat registry hives as one part of a corroborated record, not a shortcut to intent. Registry keys can look solemn while quietly refusing to name the guilty value.

===== Use Notes =====
This article is for defensive education and technical reference. It should not be treated as legal, forensic, investigative, compliance, or operational advice without qualified professional judgment.