Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
3,324 detections
Filters
Last updated
All Time
Detection languages
1,225
935
481
239
184
Contributors
626
512
469
337
261
Categories
20,443
11,831
5,875
5,151
4,864
3,324
Platforms
1,411
1,220
873
142
138
Products / Services
973
444
393
392
360
MITRE Techniques
1,525
1,085
885
593
540
CVEs
42
28
26
26
24
IDS Classtypes
390
220
127
40
26
IDS Protocols
671
150
30
18
12
Detects a multi-stage attack pattern against publicly exposed AI service endpoints. The detection correlates: (1) Enumeration of sensitive AI/MCP-related API endpoints (minimum 3 distinct hits) in a 10-minute window; (2) Subsequent POST requests to AI service endpoints within 15 minutes of the enumeration; and (3) A spike of 5 or more server errors (HTTP 5xx) within 10 minutes following the POST activity. This pattern suggests automated reconnaissance followed by attempted exploitation of exposed AI model interfaces or orchestration services. Covers T1595.002, T1190
Detects a multi-stage attack pattern against publicly exposed AI service endpoints. The detection correlates: (1) Enumeration of sensitive AI/MCP-related API endpoints (minimum 3 distinct hits) in a 10-minute window; (2) Subsequent POST requests to AI service endpoints within 15 minutes of the enumeration; and (3) A spike of 5 or more server errors (HTTP 5xx) within 10 minutes following the POST activity. This pattern suggests automated reconnaissance followed by attempted exploitation of exposed AI model interfaces or orchestration services. Covers T1595.002, T1190
Detects a multi-stage attack pattern against publicly exposed AI service endpoints. The detection correlates: (1) Enumeration of sensitive AI/MCP-related API endpoints (minimum 3 distinct hits) in a 10-minute window; (2) Subsequent POST requests to AI service endpoints within 15 minutes of the enumeration; and (3) A spike of 5 or more server errors (HTTP 5xx) within 10 minutes following the POST activity. This pattern suggests automated reconnaissance followed by attempted exploitation of exposed AI model interfaces or orchestration services. Covers T1595.002, T1190
Detects the suspicious registration of a Cloud Filter provider callback using staging paths and naming conventions associated with the ShieldCrash (CVE-2026-69414) exploit. The rule monitors DeviceEvents for cloud provider activity that originates from unsigned or untrusted binaries, excluding known-legitimate cloud synchronization clients.
Detects malformed HEIF/HEIC image files with anomalous ftyp/box structure consistent with the libheif memory corruption exploit (CVE-2026-32882) used against Discourse forum image upload endpoints
Detects malformed HEIF/HEIC image files with anomalous ftyp/box structure consistent with the libheif memory corruption exploit (CVE-2026-32882) used against Discourse forum image upload endpoints
Detects high-volume POST requests containing HEIF, HEIC, or AVIF image content types, indicative of automated fuzzing or exploitation attempts against an image-processing endpoint targeting CVE-2026-32882.
Detects high-volume POST requests containing HEIF, HEIC, or AVIF image content types, indicative of automated fuzzing or exploitation attempts against an image-processing endpoint targeting CVE-2026-32882.
Detects high-volume POST requests containing HEIF, HEIC, or AVIF image content types, indicative of automated fuzzing or exploitation attempts against an image-processing endpoint targeting CVE-2026-32882.
Detects rapid AI agent state transitions (from 'think' to 'act') within Microsoft Edge processes, specifically identifying potential exploitation of the CVE-2026-55945 race condition vulnerability, which could be used to force the unauthorized execution of an injected prompt.
Detects rapid AI agent state transitions (from 'think' to 'act') within Microsoft Edge processes, specifically identifying potential exploitation of the CVE-2026-55945 race condition vulnerability, which could be used to force the unauthorized execution of an injected prompt.
Detects unauthenticated HTTP POST requests targeting the FortiGate administrative API (/api/v2/), which may indicate exploitation attempts related to CVE-2025-25249. The rule specifically looks for requests lacking authentication headers (Cookie or Authorization).
Detects delivery and execution artifacts associated with the BlueMoon CVE-2026-85046 V8 type-confusion exploit. The rule monitors for known exploit-related file names (e.g., driver-html.js) and characteristic string indicators (e.g., v8ctf_exp_attempt, addrof, fakeobj) across file system, process execution, and network telemetry.
This rule detects malformed SCTP packets that trigger a protocol decoding error ('pkt_too_small'). This behavior is associated with the 'DiagSpill' vulnerability (CVE-2026-74469), where specially crafted chunk sequences in the SCTP stream can cause a remote service crash or potential instability. The rule includes a threshold to mitigate noise from non-malicious malformed packets.
Detects anomalous staging activity associated with the 'PPPoEject' exploit. The rule identifies the creation of GRE or IP6GRE interfaces on bonding devices immediately preceding or concurrent with FUSE filesystem mounting or I/O activity. This pattern matches the exploitation workflow of racing pppoe_sendmsg() skb reallocation via dev_hard_header() to achieve privilege escalation.
Detects a specifically malformed IPv6 Routing Header (type 0) with a 'segments_left' value of 255 preceding an Authentication Header (AH). This signature targets an exploit pattern known as 'DirtyAH6', which aims to trigger a remote kernel crash or potential remote code execution on vulnerable systems by manipulating IPv6 header processing.
Detects a specifically malformed IPv6 Routing Header (type 0) with a 'segments_left' value of 255 preceding an Authentication Header (AH). This signature targets an exploit pattern known as 'DirtyAH6', which aims to trigger a remote kernel crash or potential remote code execution on vulnerable systems by manipulating IPv6 header processing.
Detects anomalous process execution patterns involving netkit devices and virtual network interfaces (VXLAN/OVS) configured with oversized headroom, characteristic of the TUNderflow (CVE-2026-81000) exploitation chain. The rule monitors for the staging of netkit devices and specific OVS command-line interactions by non-privileged accounts, excluding known legitimate SDN and CNI infrastructure management processes.
Detects anomalous process execution patterns involving netkit devices and virtual network interfaces (VXLAN/OVS) configured with oversized headroom, characteristic of the TUNderflow (CVE-2026-81000) exploitation chain. The rule monitors for the staging of netkit devices and specific OVS command-line interactions by non-privileged accounts, excluding known legitimate SDN and CNI infrastructure management processes.
This rule detects kernel crash signatures (e.g., use-after-free, kernel panic) in the pppoe module, correlated with preceding anomalous activity related to PPPoEject or PPPoE socket manipulation. It identifies potential exploitation of CVE-2026-68121 on non-gateway devices.
Detects a suspected exploitation attempt of CVE-2026-81000, known as TUNderflow. The detection looks for a sequence of events where a process initializes a user namespace (unshare) followed by a kernel panic or error report from the tun/tap driver originating from the same process lineage within a short timeframe, which is indicative of an integer overflow exploit leading to kernel instability.
Page 24 of 159