Arnold Chan
@slazTrusted contributorCompletionist
2 followers2,100 downloads4,386 copies51 likes26,184 views
7,555 detections
Filters
Last updated
All Time
Detection languages
5,650
1,006
394
169
98
Categories
2,356
1,807
1,233
1,013
924
Platforms
4,191
1,622
1,250
676
676
Products / Services
2,820
1,003
991
720
485
MITRE Techniques
2,087
1,520
1,413
1,070
536
CVEs
38
33
24
22
21
IDS Classtypes
477
179
104
79
52
IDS Protocols
554
156
144
86
18
Detects reconnaissance commands (e.g., file listing, directory enumeration) executed as child processes of the JFrog Artifactory service. This behavior is often associated with the exploitation of Artifactory plugins or misconfigured endpoints that allow arbitrary code execution, enabling an attacker to perform discovery actions.
Detects the initialization of containerized LLM-based gateway services (e.g., litellm, ollama) followed closely in time by high-volume network traffic containing suspicious web injection payloads (e.g., SQLi, XSS, SSRF). The rule correlates process execution events for container launches with network activity originating from the same device to identify potential unauthorized or unsupervised automated exploitation.
Detects the creation of new AWS access keys or login profiles for a target identity immediately following the deletion or modification of a credential for the same identity, where the actors performing the two actions are different. This pattern indicates an adversary creating persistent access after a security incident responder or automated system has revoked a previously compromised credential.
Detects anomalous clipboard activity where content is repeatedly replaced by cryptocurrency wallet address patterns (Bitcoin or Ethereum) within a short timeframe. This behavior is indicative of clipboard hijacking (clipjacking), a technique used by malware like EggJagger to substitute legitimate payment addresses with attacker-controlled addresses.
Detects instances where LiteLLM, or related Python application processes (uvicorn, python), spawn a shell or command-line utility (bash, sh, id) under the root user account. This behavior is indicative of potential container breakout, application exploitation, or privilege escalation attempts originating from a web-accessible service.
Detects requests to a LiteLLM proxy that have been assigned the 'PROXY_ADMIN' role despite the absence of valid authentication credentials (Authorization headers, JWT tokens, or OAuth2 tokens). This pattern typically indicates that the LiteLLM instance is misconfigured, specifically that a 'master_key' is missing, causing the proxy to default to a permissive mode where every request is treated as having administrative privileges.
Detects instances where a LiteLLM proxy is utilized to invoke sensitive Model Context Protocol (MCP) tools (e.g., database, github, slack, jira, ci/cd) while authentication is bypassed ('allow_all_keys=true') and no explicit API key is provided. This indicates a potential unauthorized access attempt to sensitive integrations via a misconfigured proxy.
Detects the use of the default, hardcoded 'sk-1234' secret key for HS256 JWT signing in LiteLLM proxy logs. This configuration is insecure and allows for the forgery of arbitrary user session tokens.
Detects the creation or update of LiteLLM custom code guardrails that include suspicious Python snippets, specifically 'import os' and 'os.popen', which are indicative of attempts to execute arbitrary OS commands via guardrail logic (CVE-2026-59821).
Detects a suspicious sequence of events within JFrog Artifactory where an account accesses sensitive code-signing material (rubygems/gemspec/signing keys) and subsequently performs administrative actions (user addition/role assignment/privilege escalation) within a one-hour window, indicating potential account takeover or exploitation.
Page 446 of 756