Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Investigate Active Directory compromise by analyzing authentication logs, replication metadata, Group Policy changes, and Kerberos ticket anomalies to identify attacker persistence and lateral movement paths.
.claude/skills/performing-active-directory-compromise-investigation/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | — | — |
| case-08 | ✗→✓ | ▲ Improved | — | — |
| case-19 | ✗→✗ | = Same ✗ | — | — |
| case-06 | ✗→✗ | = Same ✗ | — | — |
| case-20 | ✗→✗ | = Same ✗ | — | — |
Active Directory (AD) compromise investigation is a critical incident response capability that focuses on identifying how attackers gained access to domain services, what persistence mechanisms they established, and the scope of credential compromise. Since 88% of breaches involve compromised credentials (Verizon 2025 DBIR), AD is the primary target for enterprise-wide attacks. Investigators must analyze NTDS.dit database integrity, Kerberos ticket-granting activity, Group Policy modifications, replication metadata, and privileged group membership changes to reconstruct the attack chain and determine full compromise scope.
The NTDS.dit file is the core Active Directory credential database containing all password hashes for domain accounts. Attackers commonly exfiltrate this file using tools like ntdsutil, secretsdump.py, or DCSync attacks via Mimikatz.
Detection indicators:
Golden Ticket indicators:
Silver Ticket indicators:
Kerberoasting indicators:
Track modifications to these critical groups:
1. Identify potentially compromised domain controllers
2. Collect Security, System, Directory Service event logs
3. Extract AD replication metadata using repadmin
4. Capture ntdsutil snapshots for offline analysis
5. Collect DNS server logs and zone transfer records
6. Export Group Policy Object configurations
7. Document current privileged group memberships1. Parse Event ID 4624/4625 for logon patterns
2. Identify pass-the-hash indicators (Event ID 4624 Type 3 with NTLM)
3. Analyze Event ID 4768/4769/4771 for Kerberos anomalies
4. Review Event ID 4776 for NTLM authentication failures
5. Cross-reference logon events with known compromised accounts
6. Map lateral movement paths through authentication chains1. Enumerate AdminSDHolder ACL modifications
2. Check for SID History abuse on accounts
3. Verify krbtgt account password age
4. Audit DSRM password configuration
5. Check for skeleton key malware indicators
6. Review AD Certificate Services for rogue certificates
7. Validate DNS records for poisoning1. Double-rotate krbtgt password (wait replication between rotations)
2. Reset all compromised account passwords
3. Remove unauthorized privileged group members
4. Revoke rogue certificates if AD CS compromised
5. Rebuild domain controllers from clean media if needed
6. Implement tiered administration model
7. Enable Protected Users group for privileged accounts| Event ID | Source | Description | |----------|--------|-------------| | 4624 | Security | Successful logon | | 4625 | Security | Failed logon | | 4648 | Security | Explicit credential logon | | 4662 | Security | Operation on AD object | | 4768 | Security | Kerberos TGT requested | | 4769 | Security | Kerberos service ticket requested | | 4771 | Security | Kerberos pre-authentication failed | | 4776 | Security | NTLM credential validation | | 5136 | Security | Directory object modified | | 5137 | Security | Directory object created | | 4706 | Security | Trust created | | 4707 | Security | Trust removed | | 4742 | Security | Computer account changed | | 8222 | System | Shadow copy created |
| Tool | Purpose | |------|---------| | BloodHound | Attack path mapping and privilege escalation analysis | | Pingcastle | AD security assessment and risk scoring | | Purple Knight | AD vulnerability scanning by Semperis | | ADRecon | Active Directory data gathering | | Mimikatz | Credential extraction and Kerberos analysis | | Impacket | DCSync detection and NTLM relay analysis | | Velociraptor | Remote forensic artifact collection | | Timeline Explorer | Event log timeline analysis |
| Technique | ID | Relevance | |-----------|----|-----------| | DCSync | T1003.006 | NTDS.dit credential extraction | | Golden Ticket | T1558.001 | Kerberos TGT forgery | | Silver Ticket | T1558.002 | Service ticket forgery | | Kerberoasting | T1558.003 | Service account hash extraction | | Pass-the-Hash | T1550.002 | NTLM hash reuse | | Group Policy Modification | T1484.001 | Persistence via GPO | | Account Manipulation | T1098 | Privileged group changes | | SID-History Injection | T1134.005 | Privilege escalation |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-19 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-23 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +9 percentage points is the difference between those two pass rates over the 23 comparable cases.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.