Custom Checks
Overview
You can write custom checks in Rego.
Once you finish writing custom checks, you can pass the policy files or the directory where those policies are stored with --policy
option.
trivy conf --policy /path/to/policy.rego --policy /path/to/custom_policies --namespaces user /path/to/config_dir
As for --namespaces
option, the detail is described as below.
File formats
If a file name matches the following file patterns, Trivy will parse the file and pass it as input to your Rego policy.
File format | File pattern |
---|---|
JSON | *.json |
YAML | *.yaml and *.yml |
Dockerfile | Dockerfile , Dockerfile.* , and *.Dockerfile |
Containerfile | Containerfile , Containerfile.* , and *.Containerfile |
Terraform | *.tf and *.tf.json |
Configuration languages
In the above general file formats, Trivy automatically identifies the following types of configuration files:
- CloudFormation (JSON/YAML)
- Kubernetes (JSON/YAML)
- Helm (YAML)
- Terraform Plan (JSON/Snapshot)
This is useful for filtering inputs, as described below.
Rego format
A single package must contain only one policy.
Example
# METADATA
# title: Deployment not allowed
# description: Deployments are not allowed because of some reasons.
# schemas:
# - input: schema["kubernetes"]
# custom:
# id: ID001
# severity: LOW
# input:
# selector:
# - type: kubernetes
package user.kubernetes.ID001
deny[res] {
input.kind == "Deployment"
msg := sprintf("Found deployment '%s' but deployments are not allowed", [input.metadata.name])
res := result.new(msg, input.kind)
}
In this example, ID001 "Deployment not allowed" is defined under user.kubernetes.ID001
.
If you add a new custom policy, it must be defined under a new package like user.kubernetes.ID002
.
Policy structure
# METADATA
(optional unless the check will be contributed into Trivy)-
- SHOULD be defined for clarity since these values will be displayed in the scan results
custom.input
SHOULD be set to indicate the input type the policy should be applied to. See list of available types
package
(required)-
- MUST follow the Rego's specification
- MUST be unique per policy
- SHOULD include policy id for uniqueness
- MAY include the group name such as
kubernetes
for clarity- Group name has no effect on policy evaluation
deny
(required)-
- SHOULD be
deny
or start withdeny_
- Although
warn
,warn_*
,violation
,violation_
also work for compatibility,deny
is recommended as severity can be defined in__rego_metadata__
.
- Although
- SHOULD return ONE OF:
- The result of a call to
result.new(msg, cause)
. Themsg
is astring
describing the issue occurrence, and thecause
is the property/object where the issue occurred. Providing this allows Trivy to ascertain line numbers and highlight code in the output. - A
string
denoting the detected issue- Although
object
withmsg
field is accepted, other fields are dropped andstring
is recommended ifresult.new()
is not utilised. - e.g.
{"msg": "deny message", "details": "something"}
- Although
- The result of a call to
- SHOULD be
Package
A package name must be unique per policy.
Example
package user.kubernetes.ID001
By default, only builtin.*
packages will be evaluated.
If you define custom packages, you have to specify the package prefix via --namespaces
option. By default, Trivy only runs in its own namespace, unless specified by the user. Note that the custom namespace does not have to be user
as in this example. It could be anything user-defined.
trivy conf --policy /path/to/custom_policies --namespaces user /path/to/config_dir
In this case, user.*
will be evaluated.
Any package prefixes such as main
and user
are allowed.
Metadata
The check must contain a Rego Metadata section. Trivy uses standard rego metadata to define the new policy and general information about it.
Trivy supports extra fields in the custom
section as described below.
Example
# METADATA
# title: Deployment not allowed
# description: Deployments are not allowed because of some reasons.
# custom:
# id: ID001
# severity: LOW
# input:
# selector:
# - type: kubernetes
If you are creating checks for your Trivy misconfiguration scans, some fields are optional as referenced in the table below. The schemas
field should be used to enable policy validation using a built-in schema. It is recommended to use this to ensure your checks are
correct and do not reference incorrect properties/values.
Field name | Allowed values | Default value | In table | In JSON |
---|---|---|---|---|
title | Any characters | N/A | ||
description | Any characters | |||
schemas.input | schema["kubernetes"] , schema["dockerfile"] , schema["cloud"] |
(applied to all input types) | ||
custom.id | Any characters | N/A | ||
custom.severity | LOW , MEDIUM , HIGH , CRITICAL |
UNKNOWN | ||
custom.recommended_actions | Any characters | |||
custom.deprecated | true , false |
false |
||
custom.input.selector.type | Any item(s) in this list | |||
url | Any characters |
custom.avd_id and custom.id
The AVD_ID can be used to link the check to the Aqua Vulnerability Database (AVD) entry. For example, the avd_id
AVD-AWS-0176
is the ID of the check in the AWS Vulnerability Database. If you are contributing your check to trivy-policies, you need to generate an ID using make id
in the trivy-checks repository. The output of the command will provide you the next free IDs for the different providers in Trivy.
The ID is based on the AVD_ID. For instance if the avd_id
is AVD-AWS-0176
, the ID is ID0176
.
custom.provider
The provider
field references the provider available in Trivy. This should be the same as the provider name in the pkg/iac/providers
directory, e.g. aws
.
custom.service
Services are defined within a provider. For instance, RDS is a service and AWS is a provider. This should be the same as the service name in one of the provider directories. (Link), e.g. aws/rds
.
custom.input
The input
tells Trivy what inputs this check should be applied to. Cloud provider checks should always use the selector
input, and should always use the type
selector with cloud
. Check targeting Kubernetes yaml can use kubenetes
, RBAC can use rbac
, and so on.
Subtypes in the custom data
Subtypes currently only need to be defined for cloud providers as detailed in the documentation.
Scan Result
Some fields are displayed in scan results.
k.yaml (kubernetes)
───────────────────
Tests: 32 (SUCCESSES: 31, FAILURES: 1, EXCEPTIONS: 0)
Failures: 1 (UNKNOWN: 0, LOW: 1, MEDIUM: 0, HIGH: 0, CRITICAL: 0)
LOW: Found deployment 'my-deployment' but deployments are not allowed
════════════════════════════════════════════════════════════════════════
Deployments are not allowed because of some reasons.
────────────────────────────────────────────────────────────────────────
k.yaml:1-2
────────────────────────────────────────────────────────────────────────
1 ┌ apiVersion: v1
2 └ kind: Deployment
────────────────────────────────────────────────────────────────────────
Input
You can specify input format via the custom.input
annotation.
Example
# METADATA
# custom:
# input:
# combine: false
# selector:
# - type: kubernetes
combine
(boolean)- The details are here.
selector
(array)-
This option filters the input by file format or configuration language. In the above example, Trivy passes only Kubernetes files to this policy. Even if a Dockerfile exists in the specified directory, it will not be passed to the policy as input.
Possible values for input types are:
dockerfile
(Dockerfile)kubernetes
(Kubernetes YAML/JSON)rbac
(Kubernetes RBAC YAML/JSON)cloud
(Cloud format, as defined by Trivy - this is used for Terraform, CloudFormation, and Cloud/AWS scanning)yaml
(Generic YAML)json
(Generic JSON)toml
(Generic TOML)
When configuration languages such as Kubernetes are not identified, file formats such as JSON will be used as
type
. When a configuration language is identified, it will overwritetype
.Example
pod.yaml
including Kubernetes Pod will be handled askubernetes
, notyaml
.type
is overwritten bykubernetes
fromyaml
.type
acceptskubernetes
,dockerfile
,cloudformation
,terraform
,terraformplan
,json
, oryaml
.
Schemas
See here for the detail.