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.
====== EDR Telemetry ======

===== Definition =====
EDR Telemetry is a log or timeline concept used to organize events, test sequence, and decide how much confidence a chronology deserves.
In a forensic report, edr telemetry 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 =====
Timelines are useful until someone treats them like prophecy. Logs can be late, clocks can drift, and duplicate timestamps can stand in a row looking innocent.
The practical job is to preserve context before the report starts turning fragments into biography. A timeline is a map, not a verdict. Maps still need scale, legend, and someone awake at the wheel.

===== Technical Description =====
Analysis compares event sources, record IDs, clock settings, timezone handling, collection timing, retention behavior, and independent artifact support.
Version, platform, retention, and collection scope should be documented before the artifact is promoted to a finding.

===== Forensic Relevance =====
  * **Timeline reconstruction:** EDR Telemetry helps test sequence, event order, and whether timestamps agree with the surrounding record.
  * **Gap analysis:** It can distinguish suspicious gaps from retention, rollover, disabled logging, or collection limits.
  * **Contradiction testing:** It is useful when logs, metadata, endpoint telemetry, and account activity tell different stories.
  * **Attribution limits:** It can support account or host context without proving who was at the keyboard.
  * **Reporting decisions:** It helps keep conclusions about edr telemetry proportional to the number and quality of sources.

===== Evidence Sources =====
^ Evidence source ^ What it may show ^ Reliability limits ^ What it cannot prove alone ^
| Event records | Show discrete actions, status changes, failures, or detections. | Provider behavior and configuration define what gets logged. | Do not prove the whole event sequence. |
| Clock and timezone settings | Help normalize times across systems. | Manual changes, drift, and daylight saving rules can distort comparison. | Do not make all timestamps equally reliable. |
| Retention and rotation settings | Explain missing or truncated logs. | Can look suspicious even when policy caused the gap. | Do not prove tampering by themselves. |
| Cross-source timelines | Show whether independent systems agree. | Different systems record different parts of the event. | Do not remove interpretation limits. |

===== Interpretation Limits =====
The most common error is treating edr telemetry 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.
Timestamps have alibis. Some are good. Some were written by clocks that should not be allowed near evidence.

===== Common Misinterpretations =====
  * Treating a timestamp as exact truth without checking timezone, clock drift, and source semantics.
  * Ignoring benign explanations because the suspicious explanation is easier to write.
  * Treating a timeline as a verdict instead of a comparison tool.
  * Treating log gaps as malicious before testing rollover, retention, and collection delay.
  * Mixing local time, UTC, and normalized time without labeling the mess.

===== Example Scenario =====
An examiner finds event records related to edr telemetry during a triage review.
The record may be meaningful, but it has to be compared with clock and timezone settings before the report gets confident.
The report should state the observed record, the collection limits, and the alternative explanations that were tested.

===== Analysis Workflow =====
  - Define the question before opening another parser: what should EDR Telemetry help answer?
  - Preserve the source evidence and document how the event records 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 edr telemetry 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 event records are consistent with activity related to edr telemetry, 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 | Event records exist, but collection scope, time source, or surrounding context is limited. | State the observation and keep interpretation narrow. |
| Moderate | Event records align with Clock and timezone settings, 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:logs_timelines:account_logon_events|Account Logon Events]] - Compare this concept when testing sequence and corroboration.
  * [[wiki:digital_forensics:logs_timelines:authentication_logs|Authentication Logs]] - Helps keep the finding grounded in a wider evidence pattern.
  * [[wiki:digital_forensics:logs_timelines:boot_time_artifacts|Boot Time Artifacts]] - Helps keep the finding grounded in a wider evidence pattern.
  * [[wiki:digital_forensics:logs_timelines:clock_drift|Clock Drift]] - Helps keep the finding grounded in a wider evidence pattern.
  * [[wiki:digital_forensics:logs_timelines:cloud_audit_timelines|Cloud Audit Timelines]] - Helps keep the finding grounded in a wider evidence pattern.
  * [[wiki:digital_forensics:logs_timelines:collection_time_effects|Collection Time Effects]] - Useful when checking whether another artifact supports the same interpretation.

===== Tools =====
  * [[https://github.com/log2timeline/plaso|Plaso]] - Timeline extraction and normalization across many artifact types.
  * [[https://github.com/google/timesketch|Timesketch]] - Collaborative timeline analysis and event tagging.
  * [[https://github.com/log2timeline/dfdatetime|dfDateTime]] - Timestamp conversion and interpretation support.
  * [[https://gchq.github.io/CyberChef/|CyberChef]] - Timestamp conversion, decoding, and quick data transformations.
  * [[https://jqlang.org/|jq]] - JSON parsing for logs, exports, and cloud evidence.
  * [[https://github.com/Yamato-Security/hayabusa|Hayabusa]] - Windows event log hunting and timeline support.
  * [[https://github.com/mikefarah/yq|yq]] - YAML, JSON, XML, and properties parsing during evidence review.
  * [[https://github.com/log2timeline/dfwinreg|dfWinReg]] - Windows Registry parsing support library.
  * [[https://ericzimmerman.github.io/#!index.md|Eric Zimmerman's Tools]] - Windows artifact parsers including MFTECmd, PECmd, EvtxECmd, and others.
  * [[https://www.sans.org/tools/kape|KAPE]] - Targeted artifact collection and processing.
  * [[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://plaso.readthedocs.io/|Plaso documentation]] - Timeline generation documentation.
  * [[https://learn.microsoft.com/en-us/windows/win32/eventlog/event-logging|Microsoft Windows Event Log]] - Event logging reference.
  * [[https://csrc.nist.gov/publications/detail/sp/800-92/final|NIST SP 800-92]] - Guide to computer security log management.
  * [[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:logs_timelines:start|Logs Timelines]]
  * [[wiki:digital_forensics:logs_timelines:account_logon_events|Account Logon Events]]
  * [[wiki:digital_forensics:logs_timelines:authentication_logs|Authentication Logs]]
  * [[wiki:digital_forensics:logs_timelines:boot_time_artifacts|Boot Time Artifacts]]
  * [[wiki:digital_forensics:logs_timelines:clock_drift|Clock Drift]]
  * [[wiki:digital_forensics:logs_timelines:cloud_audit_timelines|Cloud Audit Timelines]]
  * [[wiki:digital_forensics:logs_timelines:collection_time_effects|Collection Time Effects]]
  * [[wiki:digital_forensics:logs_timelines:cross_host_correlation|Cross-Host Correlation]]
  * [[wiki:digital_forensics:logs_timelines:dns_logs|DNS Logs]]
  * [[wiki:digital_forensics:logs_timelines:duplicate_timestamps|Duplicate Timestamps]]
  * [[wiki:digital_forensics:logs_timelines:event_record_ids|Event Record IDs]]
  * [[wiki:tools:start|Forensic Tools]]
  * [[wiki:resources:start|Project Resources]]

===== Reader Takeaway =====
EDR Telemetry is useful when it helps explain what the evidence can support and where the limits begin. Treat edr telemetry as one part of a corroborated record, not a shortcut to intent. Timestamps have alibis. Some are good. Some were written by clocks that should not be allowed near evidence.

===== 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.