Detections
Explore public detection logic contributed by the community across SIEM and rule languages.
411 detections
Filters
Last updated
All Time
Detection languages
181
74
63
61
11
Contributors
113
62
55
52
32
Categories
140
133
102
96
49
Platforms
309
74
43
15
15
Products / Services
10,665
9,758
6,692
4,421
3,749
411
MITRE Techniques
126
125
81
72
64
CVEs
5
1
1
1
1
Detects the use of the 'watch' command in conjunction with 'cat' to continuously poll the GitHub CLI config file (.config/gh/hosts.yml). This activity could indicate an attacker attempting to capture new GitHub tokens as soon as a developer logs in.
This rule detects an unusually high volume of operations related to GitHub Enterprise repositories, such as adding users, creating teams, pushing to repositories, or changing permissions, within a short timeframe. This could indicate a compromised account or automated malicious activity targeting source code repositories.
Detects suspicious process execution patterns associated with the Miasma worm, which attempts to scrape GitHub Actions Runner.Worker process memory to extract masked secrets. This rule looks for specific command-line arguments used by memory dumping tools or scripts targeting the GitHub Actions Runner.Worker process, including `grep` for secret JSON structures, `python` targeting `/proc/pid/mem`, and specific shell pipeline combinations observed in Miasma payloads.
Detects suspicious commits pretending to be the Claude AI bot but pushing dependency updates, a tactic Miasma uses to blend in with legitimate automated pull requests.
Detects automated API calls listing secrets via GitHub Actions with high pagination, indicating a script enumerating organizational secrets.
Detects unauthorized reading of the GitHub CLI hosts.yml file by processes other than standard development tools (gh, gh.exe, code). This activity could indicate an attempt to steal authenticated GitHub sessions, as the hosts.yml file contains sensitive GitHub authentication tokens.
This rule detects the creation of a new GitHub repository from a compromised endpoint, specifically looking for indicators of stolen credential exfiltration. The rule identifies GitHub application events where a repository is created and the raw event data contains keywords associated with credential dumping or specific campaign names like "Miasma" or "Spreading Blight". Automated exfiltration to GitHub is often fast and difficult to remediate, making prompt detection crucial. If triggered, it's recommended to revoke tokens used to create the repository and investigate the endpoint for compromise.
This rule detects misconfigurations in GitHub Actions workflows related to Claude automation. Specifically, it looks for workflow files (.yml or .yaml) within the '.github/workflows' directory where the initiating process command line contains 'allowed_non_write_users' or 'claude-code-action' along with a wildcard ('*') in the user permissions. Such a configuration could allow unauthorized users to trigger code changes via AI actions on a repository, indicating a potential security vulnerability.
Detects `grep` commands searching for the exact JSON structure (`"isSecret": true`) used by GitHub Actions to store masked secrets in memory. This activity could indicate an adversary attempting to discover sensitive information.
This rule looks for exposed pairs of credentials used for GitHub cloud identity federation in logs or artifacts. Leaked pairs let attackers mint tokens as your pipeline. Rotate federation settings and audit recent workflow runs.
This rule detects when a GitHub workflow branch is deleted shortly after build artifacts were downloaded. This behavior can indicate an attacker attempting to hide their tracks after exfiltrating secrets or sensitive information from build artifacts. It is crucial to preserve logs and investigate who deleted the branch. Further investigation should include questioning the user about recent software installations or clicked links, and involving the security team if the answers are unclear.
Detects the command line sequence used by Miasma to restore passwordless sudo for the 'runner' user in GitHub Actions by overwriting sudoers via a Docker container. This specific sequence involves echoing 'runner ALL=(ALL) NOPASSWD:ALL' into a file within the /mnt/runner directory and then changing its permissions to 0440, indicative of a privileged modification to the sudoers file within a containerized GitHub Actions runner environment.
Detects Python scripts accessing the /proc/*/mem file of the Runner.Worker process. This activity is indicative of memory scraping, potentially to extract sensitive information like masked GitHub Actions secrets, as observed with malware like Miasma.
Detects known joke or malicious commit messages previously observed from the Miasma threat actor when manipulating repositories. The rule specifically looks for the strings "OhNoWhatsGoingOnWithGitHub", "DontRevokeOrItGoesBoom", "thebeautifulmarchoftime", and "firedalazer" within Git commit messages.
This rule detects commands executed within GitHub Actions build jobs that attempt to read memory from the GitHub Actions runner process. This technique is often used in supply-chain attacks to extract secrets that reside in the runner's memory during job execution. The rule specifically looks for command-line arguments containing '/proc' and 'Runner.Worker' along with keywords like 'mem', 'cmdline', 'grep', 'tr -d', or 'isSecret'.
This rule detects suspicious GitHub workflow modifications or pushes originating from multiple distinct IP addresses within a short time frame (1 hour). It specifically looks for 'WorkflowCreated', 'WorkflowUpdated', or 'Push' actions on GitHub, and triggers if more than one such event and more than one distinct IP address are observed for a given account within an hour. This could indicate compromised credentials or unauthorized access to a GitHub account being used to manipulate CI/CD pipelines.
Recent reporting indicates the campaign targeted GitHub repositories by inserting malicious GitHub Actions / CI workflows designed to steal secrets and credentials.
5,561 repositories were affected in roughly six hours
Attackers pushed 5,718 malicious commits.
The objective was credential theft from CI/CD environments, including:
cloud credentials
SSH keys
access tokens
CI/CD secrets
The campaign reportedly used:
fake automation-style commit identities such as build-bot, ci-bot, or similar
malicious .github/workflows changes
workflow triggers that either:
executed automatically on push / pull request, or
remained dormant until manually triggered
Web reporting further states the payload collected environment variables, cloud metadata credentials, and repository runner secrets, then exfiltrated them to 216.126.225.129:8443.
5,561 repositories were affected in roughly six hours
Attackers pushed 5,718 malicious commits.
The objective was credential theft from CI/CD environments, including:
cloud credentials
SSH keys
access tokens
CI/CD secrets
The campaign reportedly used:
fake automation-style commit identities such as build-bot, ci-bot, or similar
malicious .github/workflows changes
workflow triggers that either:
executed automatically on push / pull request, or
remained dormant until manually triggered
Web reporting further states the payload collected environment variables, cloud metadata credentials, and repository runner secrets, then exfiltrated them to 216.126.225.129:8443.
This rule detects potential GitHub token theft or abuse activity by identifying GitHub tokens that are used from an unusually high number of distinct IP addresses or for an unusually high number of uses within a one-hour window. Specifically, it looks for tokens with more than 5 uses and more than 2 distinct IP addresses in an hour. This behavior could indicate that a compromised token is being used by an attacker from multiple locations or in an automated fashion.
Detects suspicious attempts to dump serialized GitHub Actions secrets (JSON) into a text file named 'format-results.txt' for subsequent artifact exfiltration, a technique used by IronWorm. The rule looks for shell commands that use 'echo' to write to 'format-results.txt' and contain references to 'VARIABLE_STORE' or 'secrets', indicating an attempt to exfiltrate sensitive CI/CD environment variables.
This rule detects suspicious GitHub workflow modifications or pushes originating from multiple distinct IP addresses within a short time frame (1 hour). It specifically looks for 'WorkflowCreated', 'WorkflowUpdated', or 'Push' actions on GitHub, and triggers if more than one such event and more than one distinct IP address are observed for a given account within an hour. This could indicate compromised credentials or unauthorized access to a GitHub account being used to manipulate CI/CD pipelines.
Detects file write operations to '/mnt/runner' indicating a sudoers configuration modification. This technique is utilized by the Miasma npm malware to grant passwordless sudo access for privilege escalation on GitHub Actions CI/CD runners.
Page 11 of 20