Azure AD Graph Potential Enumeration (ROADrecon)
Detects a burst of Azure AD Graph (graph.windows.net) directory enumeration by a single identity, keyed on the internal
API version that ROADrecon hardcodes (1.61-internal, with 1.6-internal as a related tooling value) plus a per-user
request burst. ROADrecon's gather command walks the directory via this internal API version using aiohttp.
This rule also detects the "evade" argument that routes every read through a single "/$batch" OData endpoint under a benign
"Microsoft ADO.NET Data Services" user-agent. A lower burst threshold applies when a scripting user-agent (aiohttp/python/fasthttp)
is present and a higher volume threshold applies otherwise.
Elastic rule (View on GitHub)
1[metadata]
2creation_date = "2026/05/20"
3integration = ["azure"]
4maturity = "production"
5updated_date = "2026/10/09"
6
7[rule]
8author = ["Elastic"]
9description = """
10Detects a burst of Azure AD Graph (graph.windows.net) directory enumeration by a single identity, keyed on the internal
11API version that ROADrecon hardcodes (`1.61-internal`, with `1.6-internal` as a related tooling value) plus a per-user
12request burst. ROADrecon's gather command walks the directory via this internal API version using aiohttp.
13This rule also detects the "evade" argument that routes every read through a single "/$batch" OData endpoint under a benign
14"Microsoft ADO.NET Data Services" user-agent. A lower burst threshold applies when a scripting user-agent (aiohttp/python/fasthttp)
15is present and a higher volume threshold applies otherwise.
16"""
17false_positives = [
18 """
19 Legitimate first-party Microsoft components (for example the Azure Graph Client Library used by portal or Office
20 experiences) occasionally call AAD Graph with the internal API version, but at single-digit to low-double-digit
21 volume; the burst thresholds exclude them. Validate the calling `azure.aadgraphactivitylogs.properties.app_id` and
22 user if a low-volume match appears.
23 """,
24 """
25 Developer activity using aiohttp or python HTTP libraries against AAD Graph for prototyping. Rare in production
26 tenants and typically low-volume.
27 """,
28 """
29 Authorized red team activity exercising ROADrecon. Document the engagement window and add exceptions on the source
30 IP or calling user.
31 """,
32]
33from = "now-9m"
34language = "esql"
35license = "Elastic License v2"
36name = "Azure AD Graph Potential Enumeration (ROADrecon)"
37note = """## Triage and analysis
38
39### Investigating Azure AD Graph Potential Enumeration (ROADrecon)
40
41This is an ES|QL aggregation rule. Alert documents contain summarized fields per burst window: the calling identity, the tenant, and a one-minute bucket. The alert itself is the signal that something resembling ROADrecon's `gather` walk happened against AAD Graph; the actual investigation happens against the raw `logs-azure.aadgraphactivitylogs-*` events for the same identity and window.
42
43The rule anchors on the hardcoded internal API version and a per-user burst rather than the user-agent alone, so it catches all three `gather` transports:
44
45- Default: `aiohttp` user-agent across ~16 directory collections.
46- `--evade`: a single `/$batch` OData endpoint under a `Microsoft ADO.NET Data Services` user-agent (defeats user-agent and distinct-endpoint based matching).
47- `-ua <value>`: an arbitrary browser user-agent.
48
49### Possible investigation steps
50
51- Confirm the burst by filtering raw AAD Graph activity for the alerting user, tenant, and time window.
52 - Filter `logs-azure.aadgraphactivitylogs-*` on the alerting user, tenant, and burst window.
53 - A default `gather` walks ~16 collections (many distinct `url.path`); `--evade` concentrates volume on `/$batch` (high `request_count`, one path).
54- Tool fingerprint: the internal API version is the primary anchor.
55 - `api_version = 1.61-internal` (hardcoded in `gather.py`, returns internal-only fields like `strongAuthenticationDetail`). No first-party Microsoft component pins this at volume.
56 - `user_agent.original` is a secondary tell: `aiohttp`/`python` (overt), `Microsoft ADO.NET Data Services` with `/$batch` (the `--evade` signature), or a browser string (`-ua`).
57- Calling client + auth method: the typical device-code-flow ROADrecon entrypoint.
58 - ROADrecon is usually pointed at the Azure CLI client (`04b07795-…`) via the `-c` flag.
59 - Uses a public-client auth method (no client secret or certificate).
60- HTTP shape distinguishes enumeration from operator follow-on.
61 - `gather` reads only, so GETs dominate (the `--evade` `/$batch` requests are POSTs to the batch endpoint).
62 - A 403/404 tail indicates the identity probing endpoints it lacks permission for.
63 - PATCH / POST / DELETE against directory objects in the same burst means the operator did more than enumerate.
64- Source posture: residential ISP, generic VPS, or anonymising-network egress raises triage priority.
65- Pivot to sign-in logs (`logs-azure.signinlogs-*`) via the sign-in correlation ID on each AAD Graph event to land on the originating token-mint.
66- Pivot to audit logs (`logs-azure.auditlogs-*`) for any directory writes by the same user near the burst that suggest persistence or modification activity.
67- Confirm the activity is not attributable to authorized testing before treating as malicious.
68 - Check for red team engagement, penetration test, or internal tooling validation.
69 - Validate against the engagement window and the operator's known source range.
70
71### Response and remediation
72
73- Enumerate device registrations created by the user during or around the burst window.
74 - `GET /v1.0/users/{id}/registeredDevices` and `GET /v1.0/users/{id}/ownedDevices`.
75 - De-register anything not attributable to a known endpoint via `DELETE /v1.0/devices/{deviceObjectId}`.
76 - Do this BEFORE session revocation: device-bound PRTs survive `revokeSignInSessions`.
77- Revoke refresh tokens and active sessions for the calling user.
78 - `POST /v1.0/users/{id}/revokeSignInSessions`.
79- Temporarily disable the user if the alert is high-confidence or you need to halt further activity while investigation continues.
80 - `PATCH /v1.0/users/{id}` with body `{"accountEnabled": false}`.
81- Audit OAuth grants and app role assignments the user holds; revoke anything minted from a kit-egress or otherwise suspicious source.
82 - `GET /v1.0/oauth2PermissionGrants?$filter=principalId eq '{id}'`, revoke via `DELETE /v1.0/oauth2PermissionGrants/{grantId}`.
83 - `GET /v1.0/users/{id}/appRoleAssignments`, revoke via `DELETE /v1.0/servicePrincipals/{spId}/appRoleAssignedTo/{assignmentId}`.
84- Reset the user's password and audit authentication methods added during the window.
85 - `GET /v1.0/users/{id}/authentication/methods` to list.
86 - Remove anything unexpected via the method-type-specific endpoint.
87- Audit directory writes by the user near the burst and roll back unauthorized changes.
88 - Query `logs-azure.auditlogs-*` for `Register device`, `Update user`, `User registered security info`, role assignment activity by the same user in the window.
89- If the calling application has no legitimate AAD Graph dependency, block further use by that app.
90 - `PATCH /beta/applications/{id}` with body `{"authenticationBehaviors": {"blockAzureADGraphAccess": true}}`.
91 - This property lives on the Graph beta endpoint, not v1.0.
92"""
93references = [
94 "https://github.com/dirkjanm/ROADtools",
95 "https://github.com/dirkjanm/ROADtools/blob/master/roadrecon/roadtools/roadrecon/gather.py",
96 "https://unit42.paloaltonetworks.com/roadtools-cloud-attacks/",
97 "https://learn.microsoft.com/en-us/graph/migrate-azure-ad-graph-overview",
98]
99risk_score = 73
100rule_id = "80aa6cca-b343-457b-877e-5877cd71a1f8"
101setup = """#### Azure AD Graph Activity Logs
102Requires Azure AD Graph Activity Logs ingested into `logs-azure.aadgraphactivitylogs-*` via the Elastic Azure integration. Enable the `AzureADGraphActivityLogs` diagnostic-settings category on Entra ID.
103"""
104severity = "high"
105tags = [
106 "Domain: Cloud",
107 "Data Source: Azure",
108 "Data Source: Azure AD Graph",
109 "Data Source: Azure AD Graph Activity Logs",
110 "Use Case: Threat Detection",
111 "Tactic: Discovery",
112 "Resources: Investigation Guide",
113 "Noise: Unknown",
114 "Performance: Fast",
115 "Rule Type: ES|QL",
116 "Platform: Entra ID",
117 "Domain: Identity",
118]
119timestamp_override = "event.ingested"
120type = "esql"
121
122query = '''
123from logs-azure.aadgraphactivitylogs-* metadata _id, _version, _index
124
125| where data_stream.dataset == "azure.aadgraphactivitylogs"
126 and azure.aadgraphactivitylogs.properties.api_version in ("1.61-internal", "1.6-internal")
127
128| eval Esql.ua_lower = to_lower(user_agent.original)
129| eval Esql.scripting_user_agent = Esql.ua_lower like "*aiohttp*" or Esql.ua_lower like "python*" or Esql.ua_lower like "*fasthttp*"
130| eval Esql.target_endpoint = case(
131 url.path like "*$batch*", "batch",
132 url.path like "*/eligibleRoleAssignments*", "eligibleRoleAssignments",
133 url.path like "*/roleAssignments*", "roleAssignments",
134 url.path like "*/users*", "users",
135 url.path like "*/groups*", "groups",
136 url.path like "*/servicePrincipals*", "servicePrincipals",
137 url.path like "*/applications*", "applications",
138 url.path like "*/devices*", "devices",
139 url.path like "*/directoryRoles*", "directoryRoles",
140 url.path like "*/roleDefinitions*", "roleDefinitions",
141 url.path like "*/administrativeUnits*", "administrativeUnits",
142 url.path like "*/contacts*", "contacts",
143 url.path like "*/oauth2PermissionGrants*", "oauth2PermissionGrants",
144 url.path like "*/authorizationPolicy*", "authorizationPolicy",
145 url.path like "*/settings*", "settings",
146 url.path like "*/policies*", "policies",
147 url.path like "*/tenantDetails*", "tenantDetails",
148 "other"
149 )
150
151| eval Esql.time_window = date_trunc(1 minutes, @timestamp)
152
153| stats
154 Esql.request_count = count(*),
155 Esql.distinct_endpoints = count_distinct(url.path),
156 Esql.collections = values(Esql.target_endpoint),
157 Esql.scripting_user_agent_seen = max(case(Esql.scripting_user_agent, 1, 0)),
158 Esql.api_versions = values(azure.aadgraphactivitylogs.properties.api_version),
159 Esql.app_ids = values(azure.aadgraphactivitylogs.properties.app_id),
160 Esql.user_agents = values(user_agent.original),
161 Esql.http_methods = values(http.request.method),
162 Esql.status_codes = values(http.response.status_code),
163 Esql.source_ips = values(source.ip),
164 Esql.source_asn_orgs = values(source.`as`.organization.name),
165 Esql.source_countries = values(source.geo.country_name),
166 Esql.actor_types = values(azure.aadgraphactivitylogs.properties.actor_type),
167 Esql.client_auth_methods = values(azure.aadgraphactivitylogs.properties.client_auth_method),
168 Esql.session_ids = values(azure.aadgraphactivitylogs.properties.session_id),
169 Esql.sign_in_activity_ids = values(azure.aadgraphactivitylogs.properties.sign_in_activity_id),
170 Esql.scopes = values(azure.aadgraphactivitylogs.properties.scopes),
171 Esql.first_seen = min(@timestamp),
172 Esql.last_seen = max(@timestamp)
173 by
174 user.id,
175 azure.tenant_id,
176 Esql.time_window
177
178| eval Esql.enum_detected = case(
179 Esql.scripting_user_agent_seen == 1 and Esql.distinct_endpoints >= 5, true,
180 Esql.distinct_endpoints >= 15, true,
181 Esql.request_count >= 150, true,
182 false
183 )
184| where Esql.enum_detected == true
185
186| keep
187 user.id,
188 azure.tenant_id,
189 Esql.*
190'''
191
192[rule.alert_suppression]
193group_by = ["user.id", "azure.tenant_id"]
194missing_fields_strategy = "suppress"
195
196[rule.alert_suppression.duration]
197unit = "m"
198value = 5
199
200[[rule.threat]]
201framework = "MITRE ATT&CK"
202[[rule.threat.technique]]
203id = "T1069"
204name = "Permission Groups Discovery"
205reference = "https://attack.mitre.org/techniques/T1069/"
206[[rule.threat.technique.subtechnique]]
207id = "T1069.003"
208name = "Cloud Groups"
209reference = "https://attack.mitre.org/techniques/T1069/003/"
210
211
212[[rule.threat.technique]]
213id = "T1087"
214name = "Account Discovery"
215reference = "https://attack.mitre.org/techniques/T1087/"
216[[rule.threat.technique.subtechnique]]
217id = "T1087.004"
218name = "Cloud Account"
219reference = "https://attack.mitre.org/techniques/T1087/004/"
220
221
222[[rule.threat.technique]]
223id = "T1526"
224name = "Cloud Service Discovery"
225reference = "https://attack.mitre.org/techniques/T1526/"
226
227
228[rule.threat.tactic]
229id = "TA0007"
230name = "Discovery"
231reference = "https://attack.mitre.org/tactics/TA0007/"
232
233[rule.investigation_fields]
234field_names = ["user.id", "azure.tenant_id"]
Triage and analysis
Investigating Azure AD Graph Potential Enumeration (ROADrecon)
This is an ES|QL aggregation rule. Alert documents contain summarized fields per burst window: the calling identity, the tenant, and a one-minute bucket. The alert itself is the signal that something resembling ROADrecon's gather walk happened against AAD Graph; the actual investigation happens against the raw logs-azure.aadgraphactivitylogs-* events for the same identity and window.
The rule anchors on the hardcoded internal API version and a per-user burst rather than the user-agent alone, so it catches all three gather transports:
- Default:
aiohttpuser-agent across ~16 directory collections. --evade: a single/$batchOData endpoint under aMicrosoft ADO.NET Data Servicesuser-agent (defeats user-agent and distinct-endpoint based matching).-ua <value>: an arbitrary browser user-agent.
Possible investigation steps
- Confirm the burst by filtering raw AAD Graph activity for the alerting user, tenant, and time window.
- Filter
logs-azure.aadgraphactivitylogs-*on the alerting user, tenant, and burst window. - A default
gatherwalks ~16 collections (many distincturl.path);--evadeconcentrates volume on/$batch(highrequest_count, one path).
- Filter
- Tool fingerprint: the internal API version is the primary anchor.
api_version = 1.61-internal(hardcoded ingather.py, returns internal-only fields likestrongAuthenticationDetail). No first-party Microsoft component pins this at volume.user_agent.originalis a secondary tell:aiohttp/python(overt),Microsoft ADO.NET Data Serviceswith/$batch(the--evadesignature), or a browser string (-ua).
- Calling client + auth method: the typical device-code-flow ROADrecon entrypoint.
- ROADrecon is usually pointed at the Azure CLI client (
04b07795-…) via the-cflag. - Uses a public-client auth method (no client secret or certificate).
- ROADrecon is usually pointed at the Azure CLI client (
- HTTP shape distinguishes enumeration from operator follow-on.
gatherreads only, so GETs dominate (the--evade/$batchrequests are POSTs to the batch endpoint).- A 403/404 tail indicates the identity probing endpoints it lacks permission for.
- PATCH / POST / DELETE against directory objects in the same burst means the operator did more than enumerate.
- Source posture: residential ISP, generic VPS, or anonymising-network egress raises triage priority.
- Pivot to sign-in logs (
logs-azure.signinlogs-*) via the sign-in correlation ID on each AAD Graph event to land on the originating token-mint. - Pivot to audit logs (
logs-azure.auditlogs-*) for any directory writes by the same user near the burst that suggest persistence or modification activity. - Confirm the activity is not attributable to authorized testing before treating as malicious.
- Check for red team engagement, penetration test, or internal tooling validation.
- Validate against the engagement window and the operator's known source range.
Response and remediation
- Enumerate device registrations created by the user during or around the burst window.
GET /v1.0/users/{id}/registeredDevicesandGET /v1.0/users/{id}/ownedDevices.- De-register anything not attributable to a known endpoint via
DELETE /v1.0/devices/{deviceObjectId}. - Do this BEFORE session revocation: device-bound PRTs survive
revokeSignInSessions.
- Revoke refresh tokens and active sessions for the calling user.
POST /v1.0/users/{id}/revokeSignInSessions.
- Temporarily disable the user if the alert is high-confidence or you need to halt further activity while investigation continues.
PATCH /v1.0/users/{id}with body{"accountEnabled": false}.
- Audit OAuth grants and app role assignments the user holds; revoke anything minted from a kit-egress or otherwise suspicious source.
GET /v1.0/oauth2PermissionGrants?$filter=principalId eq '{id}', revoke viaDELETE /v1.0/oauth2PermissionGrants/{grantId}.GET /v1.0/users/{id}/appRoleAssignments, revoke viaDELETE /v1.0/servicePrincipals/{spId}/appRoleAssignedTo/{assignmentId}.
- Reset the user's password and audit authentication methods added during the window.
GET /v1.0/users/{id}/authentication/methodsto list.- Remove anything unexpected via the method-type-specific endpoint.
- Audit directory writes by the user near the burst and roll back unauthorized changes.
- Query
logs-azure.auditlogs-*forRegister device,Update user,User registered security info, role assignment activity by the same user in the window.
- Query
- If the calling application has no legitimate AAD Graph dependency, block further use by that app.
PATCH /beta/applications/{id}with body{"authenticationBehaviors": {"blockAzureADGraphAccess": true}}.- This property lives on the Graph beta endpoint, not v1.0.
References
Related rules
- Azure AD Graph Access with Suspicious User-Agent
- Azure AD Graph High 4xx Error Ratio from User
- Entra ID OAuth Device Code Sign-in to Azure AD Graph Enumeration
- Azure AD Graph Access with Unusual Client and User
- Azure AD Graph Access with Unusual User and ASN