Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
SharpHound is the data collector for BloodHound. It reads Active Directory objects, permissions, sessions, and group memberships, then exports them for graph analysis of attack paths.

What Is SharpHound?
SharpHound is the collector half of BloodHound. BloodHound draws the graph; SharpHound gathers the data that the graph is built from. Run inside a domain, it queries Active Directory for users, computers, groups, organisational units, group policy objects, trusts, and the access control entries attached to all of them, then writes the result as JSON files for import.
The point of the exercise is relationships. Individually, a permission that lets a helpdesk group reset passwords for a set of accounts looks reasonable. Add that one of those accounts is a local administrator on a server where a domain administrator has an active session, and you have a path from helpdesk to domain compromise that nobody designed and nobody noticed. BloodHound finds that path in one query.
| Collection method | Data gathered |
|---|---|
Group | Group membership across the domain |
ACL | Access control entries, including write and reset rights over objects |
Session | Which users have sessions on which computers |
LocalAdmin and RDP | Local group membership on domain-joined machines |
Trusts | Relationships between domains and forests |
ObjectProps | Attributes such as service principal names and pre-authentication settings |
GPOLocalGroup | Local group membership applied through group policy |
Most of this comes from ordinary LDAP queries that any authenticated user may run, which is the uncomfortable part: Active Directory is designed to be readable by its members. Session and local group data require contact with individual machines, so the DCOnly method exists for operators who want directory data without touching a single workstation.
Companion collectors extend the picture. AzureHound performs the same role against Entra ID, mapping cloud roles, applications, and consent grants, which matters now that hybrid identity is the norm.
Before graph analysis, finding a privilege escalation path through a large directory was manual and slow, and testers found whatever they happened to look at. SharpHound and BloodHound made it systematic. A tester with ordinary user rights can run collection, import the data, and ask for the shortest path from their own account to Domain Admins. In an estate that has grown for fifteen years, there is almost always one.
The same capability is now routine in criminal operations, particularly ransomware intrusions, where speed matters and the graph answers the only question the attacker has: what is the fastest route to the backups and the domain controllers. It is a core part of the post-exploitation toolkit.
Defenders should run it on their own directory, and many now do. The graph identifies the problems that a configuration review misses: users with excessive ACLs over other objects, service accounts in privileged groups, unconstrained delegation on servers, domain admins logging on to workstations, and nested group membership that nobody can follow by hand.
Each of those has a fix. Tier administrative accounts so privileged credentials never appear on user devices. Remove standing membership of privileged groups in favour of just-in-time elevation. Audit and strip delegated permissions granted during long-forgotten projects. Remove unconstrained delegation. Treat the graph as a backlog and re-run collection quarterly to see whether the shortest paths are getting longer, which is a far better metric than a count of findings. Our wider Active Directory security guidance covers the same ground.
SharpHound is noticeable if the right telemetry exists. The directory signals are the LDAP queries themselves: enumerating every user, computer, and group in sequence produces a volume and pattern that no ordinary client generates. Domain controller diagnostic logging for expensive and inefficient searches, event 1644, exposes this when it is enabled, and Microsoft Defender for Identity raises reconnaissance alerts for the same behaviour.
Host and network signals add confirmation. Session enumeration contacts many machines over SMB in a short window, which is visible in network flow data. A workstation connecting to hundreds of peers within minutes is anomalous regardless of what it is sending. Decoy objects work here too: a user account with an attractive name and no legitimate purpose, alerted on whenever it is read in bulk enumeration.
Our Infrastructure Security Testing Services include directory graph analysis so you see the same attack paths an intruder would.
SharpHound collects the data; BloodHound analyses it. SharpHound is a collector run inside the target domain that queries Active Directory and exports JSON. BloodHound is the application that imports those files into a graph database and lets an analyst query relationships, such as the shortest path from a given account to Domain Admins.
No. Most collection works with any authenticated domain account, because Active Directory is designed so that members can read directory information. Elevated rights improve coverage of local group membership on remote machines, but the core attack path data, including group membership and access control entries, is available to an ordinary user.
Yes, and increasingly they do. Running collection against your own directory reveals the privilege escalation paths an attacker would use, which configuration reviews rarely surface. Treat the output as a remediation backlog, and re-run it quarterly to measure whether the shortest paths to privileged groups are becoming longer and fewer over time.
Watch for bulk directory enumeration: large sequential LDAP queries from a workstation, visible through domain controller search logging or an identity protection product. Network data showing one host contacting hundreds of peers over SMB within minutes is another strong signal. Decoy accounts that nothing legitimate should read provide a high-confidence alert when enumeration touches them.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.