Kubernetes Static Pod Manifest File Access

[metadata]
creation_date = "2026/05/06"
integration = ["endpoint", "auditd_manager"]
maturity = "production"
updated_date = "2026/05/06"

[rule]
author = ["Elastic"]
description = """
Detects Linux process executions where shells, editors, interpreters, or file/stream utilities reference
/etc/kubernetes/manifests in process arguments. That directory holds static pod manifests read by the kubelet;
interaction via editors, downloaders, kubectl, redirection helpers (tee, dd), or scripting runtimes may indicate
staging or tampering with manifests for persistence or privileged workload placement. Pairs with file-telemetry rules
that flag direct manifest creation on container workloads.
"""
false_positives = [
    """
    Cluster provisioning (kubeadm), configuration management, or administrators editing manifests during maintenance may
    match. Baseline approved automation and interactive admin sessions on control plane nodes.
    """,
]
from = "now-9m"
index = ["auditbeat-*", "logs-auditd_manager.auditd-*", "logs-endpoint.events.process*"]
language = "kuery"
license = "Elastic License v2"
name = "Kubernetes Static Pod Manifest File Access"
note = """## Triage and analysis

### Investigating Kubernetes Static Pod Manifest File Access

Review the full command line (process.args, process.command_line), user.id, user.name, process.parent, and whether the
session was interactive. Confirm if the host is a Kubernetes node or admin jump host where manifest edits are expected.

### Possible investigation steps

- Compare activity to change windows and identity baselines; prioritize events without matching change tickets.
- Inspect subsequent process and file events on the same host for writes under /etc/kubernetes/manifests or kubelet
  restarts.
- Correlate with Kubernetes audit logs and node/agent telemetry for related compromise indicators.

### Response and remediation

- If unauthorized, restore manifests from known-good sources, isolate the host, and review cluster integrity per incident
  policy.
"""
references = [
    "https://kubernetes.io/docs/tasks/configure-pod-container/static-pod/",
    "https://attack.mitre.org/techniques/T1053/007/",
]
risk_score = 47
rule_id = "a698a653-e144-4e40-bade-35135935be45"
setup = """## Setup

Requires **Elastic Defend** and/or **Auditd Manager** process telemetry (`logs-endpoint.events.process*`,
`logs-auditd_manager.auditd-*`, `auditbeat-*`) with command-line argument capture for exec events.

### Elastic Defend
Install the Elastic Defend integration via Fleet on Linux hosts and use a policy that collects process events with
arguments.

### Auditd Manager
Deploy Auditd Manager and ensure execve (or equivalent process) auditing is enabled so `process.args` and
`process.executable` populate for monitored binaries.

See https://docs.elastic.co/integrations/auditd_manager
"""
severity = "medium"
tags = [
    "Data Source: Auditd Manager",
    "Data Source: Elastic Defend",
    "Domain: Endpoint",
    "Domain: Kubernetes",
    "Domain: Container",
    "OS: Linux",
    "Use Case: Threat Detection",
    "Tactic: Persistence",
    "Tactic: Privilege Escalation",
    "Resources: Investigation Guide",
]
timestamp_override = "event.ingested"
type = "query"
query = '''
host.os.type:linux and event.category:process and event.action:(exec or executed) and 
process.name:(
  bash or sh or dash or zsh or 
  cat or cp or mv or touch or tee or dd or
  sed or awk or 
  curl or wget or scp or
  vi or vim or nano or echo or
  busybox or
  python* or perl* or ruby* or node or lua* or
  openssl or base64 or xxd or
  .*) and 
  process.args:(*/etc/kubernetes/manifests/* and not (/etc/kubernetes/manifests/etcd* or /etc/kubernetes/manifests/kube-apiserver* or /etc/kubernetes/manifests/kube-scheduler* or /etc/kubernetes/manifests/kube-controller-manager*))
'''

[[rule.threat]]
framework = "MITRE ATT&CK"

[[rule.threat.technique]]
id = "T1053"
name = "Scheduled Task/Job"
reference = "https://attack.mitre.org/techniques/T1053/"

[[rule.threat.technique.subtechnique]]
id = "T1053.007"
name = "Container Orchestration Job"
reference = "https://attack.mitre.org/techniques/T1053/007/"

[rule.threat.tactic]
id = "TA0003"
name = "Persistence"
reference = "https://attack.mitre.org/tactics/TA0003/"

[[rule.threat]]
framework = "MITRE ATT&CK"

[[rule.threat.technique]]
id = "T1543"
name = "Create or Modify System Process"
reference = "https://attack.mitre.org/techniques/T1543/"

[[rule.threat.technique.subtechnique]]
id = "T1543.005"
name = "Container Service"
reference = "https://attack.mitre.org/techniques/T1543/005/"

[rule.threat.tactic]
id = "TA0004"
name = "Privilege Escalation"
reference = "https://attack.mitre.org/tactics/TA0004/"