Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/adscanpro/claude-ad/ad-attack-pathsgit clone --depth 1 https://github.com/ADScanPro/Claude-ADWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/adscanpro/claude-ad/ad-attack-paths)<a href="https://agentmods.dev/commands/adscanpro/claude-ad/ad-attack-paths"><img src="https://agentmods.dev/badge/commands/adscanpro/claude-ad/ad-attack-paths.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00042 | $0.00395 |
| Opus 5 | $0.00021 | $0.00198 |
| Sonnet 5 | $0.00008 | $0.00079 |
| Haiku 4.5 | $0.00004 | $0.00040 |
Grade A, and why
ad-attack-paths scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Reason attack paths to Domain Admin for $1.
Authorized engagements only. This reads the inventory produced by /ad-recon.
If no inventory exists under ./ad-engagement/, tell the operator to run
/ad-recon $1 first and stop.
Steps:
-
Read
./ad-engagement/scope.md(objective and forbidden actions) and the inventory + BloodHound CE zip from/ad-recon. -
Launch the ad-attack-planner agent with those inputs. It loads the graph into BloodHound CE, reasons which edges to chain from a low-privilege principal to the objective (ACL abuse, Kerberos, ADCS ESC1/ESC8, DCSync, coercion + relay), and checks each hop against posture using
ad-environment-constraintsandcompliance-mapping. -
The planner does not execute. It writes a plan file to
./ad-engagement/and presents the ranked path(s) with, per step: the technique-skill, the exact standard-tool command, feasibility, telemetry, rollback where state changes, and a conceptual compliance theme. -
Present the plan and stop at the confirmation gate. Ask the operator which steps are approved. Only after explicit human approval does
ad-exploit-operatorrun any step, one at a time, with its own gate before anything that modifies the directory.
Recommend the quietest path that reaches the objective, but let the human decide.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 35 lines · 42 tokens per session scan A 4063fe94bf6a
ad-attack-paths is a command published in the GitHub repository ADScanPro/Claude-AD (149 stars, last pushed 10d ago), licensed MIT. It adds 42 tokens to every session and 395 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
hunt
Active Directory offensive phase — credential attacks (spray if allowed), AS-REP and Kerberoast workflows, NTLM relay opportunities, delegation abuse paths, ACL-based escalation, AD CS template abuse, lateral movement within ROE. Usage: /hunt corp.local.
recon
Active Directory reconnaissance — DNS SRV/LDAP DC discovery, LDAP user and group enumeration, Kerberos pre-auth and AS-REP candidates, SPN discovery for Kerberoasting, SMB signing and null session policy, password policy, trust relationships, BloodHound/SharpHound collection prep. Requires ROE. Usage: /recon…
autopilot
Phased autonomous checklist for AD engagements within ROE — not unsupervised internet-wide scanning. Usage: /autopilot corp.local --normal.
chain
Build multi-hop Active Directory attack paths — combine AS-REP, Kerberoast, delegation, ACL abuse, AD CS, lateral steps into one narrative. Usage: /chain.
intel
Domain controller and Windows Server build intel — patch level, known abuse techniques relevant to version, planning prioritization. Usage: /intel corp.local.
scope
Rules of engagement — verify authorized domains, DCs, subnets, accounts, destructive actions, and exclusions before testing. Usage: /scope or /scope corp.local.