This checklist is designed to get a security team from zero to their first documented hunt in one sprint. Repeat this cycle. Build the library.


Pre-Hunt: Setup

  • Define your hypothesis — make it specific and testable
  • Identify the MITRE ATT&CK tactic/technique you’re hunting
  • Document your data sources (which logs, which SIEM, what retention window)
  • Define your scope (which assets, environments, time window)
  • Confirm data availability — run a quick validation query first

Rule: If you can’t write down your hypothesis in one sentence, you’re not ready to start.


During Hunt: Execution

  • Start with a baseline query to understand normal behavior
  • Profile volume, frequency, and statistical outliers
  • Pivot on interesting artifacts (process names, hashes, IPs, user accounts)
  • Document every query you run — copy/paste into a running log
  • Note dead ends — negative findings are valid and reusable
  • Tag anything suspicious for triage: host, timestamp, artifact

Post-Hunt: Documentation

  • Write the hunt report (Hypothesis → Scope → Queries → Findings)
  • Classify findings: True Positive / False Positive / Inconclusive
  • If True Positive: escalate per IR procedures
  • Convert best query into a detection rule (Sigma, KQL, SPL)
  • Submit detection rule to your SIEM rule library
  • Schedule the hunt for re-execution (monthly or on threat intel trigger)

Tooling Baseline

You need at minimum:

ToolPurpose
SIEM (Splunk/Sentinel/Elastic)Query execution, log search
EDR (CrowdStrike/Defender/SentinelOne)Endpoint process/network telemetry
MITRE ATT&CK NavigatorTechnique mapping and coverage tracking
SigmaVendor-agnostic detection rule format
Git repoHunt report version control

Common Beginner Mistakes

  1. Hunting without a hypothesis — aimless log searching ≠ threat hunting
  2. Not documenting queries — you’ll need to rerun them next month
  3. Stopping at “nothing found” — document it as a negative; baseline it
  4. Skipping detection rule creation — the hunt wasn’t complete if you didn’t codify what you learned
  5. No peer review — always have a second set of eyes on true positives before escalation

Template: Hunt Report Header

# Hunt Report: [Title]
**Date:** YYYY-MM-DD  
**Analyst:** [Name]  
**Tactic:** [MITRE Tactic]  
**Technique:** [T-Number — Name]  

## Hypothesis
If [condition], then [indicator] will appear in [data source].

## Scope
- Data source: [SIEM + log type]
- Timeframe: [start] to [end]
- Asset coverage: [environment description]

## Queries
[SPL/KQL/query here]

## Findings
[Description or "No indicators observed"]

## Detection Output
[Sigma rule or "No rule created — negative hunt"]