agentleFS
fatihkanfatihkan/badi/.claude/skills-vault/pentest-exploit-chain/SKILL.md

pentest-exploit-chain

Skill8 starsChanged 3 months agoMITView on GitHub

Multi-step exploit chain analysis — linking low/medium severity findings into a critical chain, stealth+impact scoring advisory. Triggers on exploit chain, attack chain, chain analysis, multi-step attack, kill chain, attack path, stealth scoring, impact analysis.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write a two-sided check for each chain step:

    line 141

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

pentest-exploit-chain

Chaining individual low-to-medium severity findings into a full compromise chain. The highest value-add of engagement intelligence — human analysis only.

Triggers

  • "can these findings be chained"
  • "draw the kill chain"
  • "low + medium = critical"
  • "attack path simulation"
  • "stealth + impact score"

Lockheed Cyber Kill Chain (7 step)

1. Reconnaissance       — OSINT, port scan
2. Weaponization        — exploit + payload combine
3. Delivery             — phishing, drive-by, USB
4. Exploitation         — vulnerability trigger
5. Installation         — persistent foothold
6. Command & Control    — beacon
7. Actions on Objectives — data exfil, destruction, ransom

MITRE ATT&CK Tactic Chain

TA0043 Reconnaissance
  -> TA0001 Initial Access
    -> TA0002 Execution
      -> TA0003 Persistence
        -> TA0004 Privilege Escalation
          -> TA0005 Defense Evasion
            -> TA0006 Credential Access
              -> TA0007 Discovery
                -> TA0008 Lateral Movement
                  -> TA0009 Collection
                    -> TA0011 Command and Control
                      -> TA0010 Exfiltration
                        -> TA0040 Impact

Chain Analysis Methodology

1. List all findings (low + medium + high)
2. For each finding: input requirement, output privilege/access
3. Match: output of finding A, input of finding B
4. Draw the path of steps (DAG)
5. Shortest path -> max impact
6. A stealth score (1-5) per step
7. Total stealth + impact -> chain rank

Stealth Scoring (1-5)

ScoreTypeDetection Likelihood
1LoudEDR/SIEM auto block/alert
2NoisyCaught when a SOC analyst reviews it in 5 min
3ModerateDefault rule insufficient, hunt query needed
4QuietOnly sophisticated hunt + threat intel
5StealthUndetectable (zero-day, signed bin)

Impact Scoring (1-5)

ScoreTypeImpact
1ReconInformation gathering (no data disclosure)
2LimitedSingle account compromise
3LocalSingle system compromise
4LateralMultiple systems / DB access
5CatastrophicDA, ransomware, mass exfil

Chain Example

# Chain — "User to Domain Admin"
Total: 6 step, Stealth: 18/30, Impact: 5/5 (DA)

## Step 1: Phishing (Initial Access)
- Finding: User awareness training missing (engagement OSINT)
- Action: Spear phish IT employee
- Stealth: 3 (email gateway may catch it, 20%)
- Impact: 2 (single user cred)
- Output: domain\jdoe / Spring2026!

## Step 2: Initial Foothold (Execution)
- Finding: VPN portal MFA missing
- Action: VPN login -> during domain join
- Stealth: 4 (MFA log signal is strong but missing)
- Impact: 2
- Output: Internal network access

## Step 3: Recon (Discovery)
- Finding: SMB anonymous null session on AD
- Action: enum4linux + RPC
- Stealth: 2 (event 4624 type 3 fires)
- Impact: 2
- Output: Domain user list + group list

## Step 4: Cred Access (Credential Access)
- Finding: Kerberoastable service account weak password
- Action: GetUserSPNs.py + hashcat 4 hr crack
- Stealth: 3 (event 4769 with 0x17 encryption)
- Impact: 3
- Output: svc-mssql / Mssql2024

## Step 5: Privesc (Privilege Escalation)
- Finding: svc-mssql member of "DB Admins"
- Action: Already privileged service acct
- Stealth: 5 (no abnormal event)
- Impact: 4
- Output: Local admin on DB-PROD

## Step 6: DA (Privilege Escalation)
- Finding: DB-PROD has unconstrained delegation
- Action: TGT capture via printer bug
- Stealth: 3 (anomalous service ticket request)
- Impact: 5
- Output: Domain Admin TGT

Compensating Controls

Write a two-sided check for each chain step:

step: 4 - Kerberoast
attack:
  technique: T1558.003
  noise: Event 4769 with encryption 0x17 (RC4)
controls:
  preventive:
    - Service account: long random password (30+ char)
    - Use AES-256 only (set msDS-SupportedEncryptionTypes)
    - Microsoft Managed Service Accounts (gMSA)
  detective:
    - Sigma rule: kerberoast_rc4_encryption.yml
    - Alert threshold: > 5 TGS-REQ in 1 hr from single user
  responsive:
    - Auto-disable account on alert
    - 24/7 SOC triage SLA

Output Template

## Exploit Chain Report — <engagement>

### Top 3 Chains (ranked by Impact / Stealth)

| # | Title | Steps | Stealth | Impact | Time |
|---|-------|-------|---------|--------|------|
| 1 | User -> DA via Kerberoast | 6 | 18 | 5 | 6 hours |
| 2 | Web SSRF -> Cloud Metadata -> RDS | 4 | 12 | 5 | 3 hours |
| 3 | Phish -> Lateral -> Ransom Sim | 5 | 10 | 5 | 8 hours |

### Recommended Defense Priority
1. Add Kerberoast detection (Sigma rule attached)
2. MFA on VPN (chain step 2 break)
3. Service account password complexity (chain step 4 break)
4. Disable unconstrained delegation (chain step 6 break)

Out-of-Scope

  • Live chain execution (advisory + writeup only)
  • 0-day discovery (n-day exploits are used)

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.