Legend
BuildImageCiStep
Build image CI action stepstring
required
Unique step identifier, used for referencing in logs and dependencies(minLen:1)
string
Human-readable step name shown in the pipeline UI
string
Why this step exists and what it does, in plain language, shown in the pipeline UI
Not used when running the step
integer | null
Maximum execution time in seconds before the step is killed(format:int32,min:1; templateable)
boolean | null
Condition that must evaluate to true for this step to run
Accepts either a boolean literal or a full template token that resolves to a boolean(templateable)
string[]
Ids of the steps this step waits for. Without
needs, a step waits for
everything listed before it. With needs, it waits only for the steps
named here, so it can start alongside earlier steps it doesn’t depend on.
An empty list starts the step as soon as its list starts.
Each id must name an action step listed earlier in the same steps list
(the top-level list or the same group). A step that names a matrix step
waits for every copy. If a step it needs fails or is skipped, this step
is skipped too.StepMatrix
Runs one copy of this step for every combination of the matrix axes.
Each copy reads its values with << matrix.<axis> >> in any templatable
field, including infrastructure. Each copy runs as a step of its own,
with an id made of the step id and its axis values in axis-name order
(test-pg15-20) and a name with its values in parentheses. A copy’s id
can’t be the id of another step or of another matrix’s copy. Step
concurrency limits copies across runs: without a key each copy has its
own slot, and an explicit key is one limit for every copy. Copies of one
run don’t wait for each other’s concurrency; use max_parallel for that.
Other steps can’t read a matrix step’s outputs or inputs, since there is
one per copy.See: StepMatrix
ConcurrencyConfig
Concurrency configuration for controlling step execution across variantsSee: ConcurrencyConfig
enum
required
Allowed values:
build:imageImageBuildConfig
required
Builder configuration — dockerfile, nixpacks, or railpack settings for image buildsSee: ImageBuildConfig(templateable)
EnvironmentVariablesValue
Environment variables available during the build — plain strings or AWS secret referencesSee: EnvironmentVariablesValue(templateable)
boolean | null
Enable verbose build output for debugging build issues
Accepts either a boolean literal or a full template token(templateable)
boolean | null
Build a SOCI-enabled image for lazy loading on AWS Fargate
When enabled, image layers are gzip-compressed (SOCI cannot index zstd)
and a SOCI index manifest v2 is generated and pushed alongside the image
Accepts either a boolean literal or a full template token(templateable)
EcrDestination[]
required
ECR destinations where the built Docker image will be pushedSee: EcrDestination(minItems:1; templateable)
StepInfrastructure
required
Execution environment configuration — required because image builds run on provisioned computeSee: StepInfrastructure(templateable)
BuildStaticCiStep
Build static CI action stepstring
required
Unique step identifier, used for referencing in logs and dependencies(minLen:1)
string
Human-readable step name shown in the pipeline UI
string
Why this step exists and what it does, in plain language, shown in the pipeline UI
Not used when running the step
integer | null
Maximum execution time in seconds before the step is killed(format:int32,min:1; templateable)
boolean | null
Condition that must evaluate to true for this step to run
Accepts either a boolean literal or a full template token that resolves to a boolean(templateable)
string[]
Ids of the steps this step waits for. Without
needs, a step waits for
everything listed before it. With needs, it waits only for the steps
named here, so it can start alongside earlier steps it doesn’t depend on.
An empty list starts the step as soon as its list starts.
Each id must name an action step listed earlier in the same steps list
(the top-level list or the same group). A step that names a matrix step
waits for every copy. If a step it needs fails or is skipped, this step
is skipped too.StepMatrix
Runs one copy of this step for every combination of the matrix axes.
Each copy reads its values with << matrix.<axis> >> in any templatable
field, including infrastructure. Each copy runs as a step of its own,
with an id made of the step id and its axis values in axis-name order
(test-pg15-20) and a name with its values in parentheses. A copy’s id
can’t be the id of another step or of another matrix’s copy. Step
concurrency limits copies across runs: without a key each copy has its
own slot, and an explicit key is one limit for every copy. Copies of one
run don’t wait for each other’s concurrency; use max_parallel for that.
Other steps can’t read a matrix step’s outputs or inputs, since there is
one per copy.See: StepMatrix
ConcurrencyConfig
Concurrency configuration for controlling step execution across variantsSee: ConcurrencyConfig
enum
required
Allowed values:
build:staticStaticBuildConfig
required
Builder configuration — dockerfile, nixpacks, or railpack settingsSee: StaticBuildConfig(templateable)
EnvironmentVariablesValue
Environment variables available during the build — plain strings or AWS secret referencesSee: EnvironmentVariablesValue(templateable)
boolean | null
Enable verbose build output for debugging build issues
Accepts either a boolean literal or a full template token(templateable)
S3Destination[]
required
S3 destinations where the built static assets will be uploadedSee: S3Destination(minItems:1; templateable)
StepInfrastructure
required
Execution environment configuration — required because static builds run on provisioned computeSee: StepInfrastructure(templateable)
CustomCommandCiStep
Custom command CI action stepstring
required
Unique step identifier, used for referencing in logs and dependencies(minLen:1)
string
Human-readable step name shown in the pipeline UI
string
Why this step exists and what it does, in plain language, shown in the pipeline UI
Not used when running the step
integer | null
Maximum execution time in seconds before the step is killed(format:int32,min:1; templateable)
boolean | null
Condition that must evaluate to true for this step to run
Accepts either a boolean literal or a full template token that resolves to a boolean(templateable)
string[]
Ids of the steps this step waits for. Without
needs, a step waits for
everything listed before it. With needs, it waits only for the steps
named here, so it can start alongside earlier steps it doesn’t depend on.
An empty list starts the step as soon as its list starts.
Each id must name an action step listed earlier in the same steps list
(the top-level list or the same group). A step that names a matrix step
waits for every copy. If a step it needs fails or is skipped, this step
is skipped too.StepMatrix
Runs one copy of this step for every combination of the matrix axes.
Each copy reads its values with << matrix.<axis> >> in any templatable
field, including infrastructure. Each copy runs as a step of its own,
with an id made of the step id and its axis values in axis-name order
(test-pg15-20) and a name with its values in parentheses. A copy’s id
can’t be the id of another step or of another matrix’s copy. Step
concurrency limits copies across runs: without a key each copy has its
own slot, and an explicit key is one limit for every copy. Copies of one
run don’t wait for each other’s concurrency; use max_parallel for that.
Other steps can’t read a matrix step’s outputs or inputs, since there is
one per copy.See: StepMatrix
SetupAction[] | null
Setup actions to install tools before the commands run, in order, before any command.
Each entry installs a tool via the Flightcontrol actions registry (ravionhq/actions).See: SetupAction(templateable)
ConcurrencyConfig
Concurrency configuration for controlling step execution across variantsSee: ConcurrencyConfig
enum
required
Allowed values:
customstring[]
Glob patterns, relative to the repository root, that gate this step.
Omit the field to run the step on every run.When set, each pipeline run compares the commit
source resolves to
with the commit of this step’s latest successful execution in the same
pipeline variant. If no changed file matches a pattern, the step is
skipped with the cause NO_WATCHED_CHANGES. A skipped or failed
execution never becomes the baseline, so a change stays pending until
the step succeeds. The step always runs when it has no earlier success,
when its configuration changed since that success, or when the
comparison is inconclusive. Patterns support *, **, ?, [abc] and
{a,b}, and match dotfiles; a pattern starting with ! excludes files.
Requires source.(minItems:1)string[]
required
Shell commands to execute in order(minItems:1; templateable)
EnvironmentVariablesValue
Environment variables available during command executionSee: EnvironmentVariablesValue(templateable)
string[]
Names of the string outputs this step publishes, readable by later steps as
<< steps.<id>.output.<name> >>. Commands publish values by appending
“name=value” lines to the file at $RAVION_OUTPUT; values for names that are
not declared here are ignored.(minItems:1)
StepInfrastructure
required
Execution environment configuration — required because custom commands run on provisioned computeSee: StepInfrastructure(templateable)
TerraformApplyCiStep
Terraform apply CI action stepstring
required
Unique step identifier, used for referencing in logs and dependencies(minLen:1)
string
Human-readable step name shown in the pipeline UI
string
Why this step exists and what it does, in plain language, shown in the pipeline UI
Not used when running the step
integer | null
Maximum execution time in seconds before the step is killed(format:int32,min:1; templateable)
boolean | null
Condition that must evaluate to true for this step to run
Accepts either a boolean literal or a full template token that resolves to a boolean(templateable)
string[]
Ids of the steps this step waits for. Without
needs, a step waits for
everything listed before it. With needs, it waits only for the steps
named here, so it can start alongside earlier steps it doesn’t depend on.
An empty list starts the step as soon as its list starts.
Each id must name an action step listed earlier in the same steps list
(the top-level list or the same group). A step that names a matrix step
waits for every copy. If a step it needs fails or is skipped, this step
is skipped too.StepMatrix
Runs one copy of this step for every combination of the matrix axes.
Each copy reads its values with << matrix.<axis> >> in any templatable
field, including infrastructure. Each copy runs as a step of its own,
with an id made of the step id and its axis values in axis-name order
(test-pg15-20) and a name with its values in parentheses. A copy’s id
can’t be the id of another step or of another matrix’s copy. Step
concurrency limits copies across runs: without a key each copy has its
own slot, and an explicit key is one limit for every copy. Copies of one
run don’t wait for each other’s concurrency; use max_parallel for that.
Other steps can’t read a matrix step’s outputs or inputs, since there is
one per copy.See: StepMatrix
ConcurrencyConfig
Concurrency configuration for controlling step execution across variantsSee: ConcurrencyConfig
enum
required
Allowed values:
terraform:applyGitSource
EnvironmentVariablesValue
Environment variables available during execution — plain strings or AWS secret referencesSee: EnvironmentVariablesValue(templateable)
enum
required
IaC tool implementation used to execute this stepAllowed values:
opentofu | terraform(templateable)string
required
Terraform/OpenTofu version to execute(minLen:1; templateable)
string | null
Stack ID used for terraform resource/execution tracking
Automatically set when triggered from a stack(templateable)
string[] | null
Integrations whose credentials are injected into the Terraform/OpenTofu process.
Each entry is an integration ID or givenId; at most one per provider.(templateable)
string
required
S3 URI of the plan file to apply — typically from a preceding plan step output (e.g. s3://bucket/path/to/plan.tfplan)(templateable)
StepInfrastructure
required
Execution environment configuration — required because Terraform apply runs on provisioned computeSee: StepInfrastructure(templateable)
TerraformPlanCiStep
Terraform plan CI action stepstring
required
Unique step identifier, used for referencing in logs and dependencies(minLen:1)
string
Human-readable step name shown in the pipeline UI
string
Why this step exists and what it does, in plain language, shown in the pipeline UI
Not used when running the step
integer | null
Maximum execution time in seconds before the step is killed(format:int32,min:1; templateable)
boolean | null
Condition that must evaluate to true for this step to run
Accepts either a boolean literal or a full template token that resolves to a boolean(templateable)
string[]
Ids of the steps this step waits for. Without
needs, a step waits for
everything listed before it. With needs, it waits only for the steps
named here, so it can start alongside earlier steps it doesn’t depend on.
An empty list starts the step as soon as its list starts.
Each id must name an action step listed earlier in the same steps list
(the top-level list or the same group). A step that names a matrix step
waits for every copy. If a step it needs fails or is skipped, this step
is skipped too.StepMatrix
Runs one copy of this step for every combination of the matrix axes.
Each copy reads its values with << matrix.<axis> >> in any templatable
field, including infrastructure. Each copy runs as a step of its own,
with an id made of the step id and its axis values in axis-name order
(test-pg15-20) and a name with its values in parentheses. A copy’s id
can’t be the id of another step or of another matrix’s copy. Step
concurrency limits copies across runs: without a key each copy has its
own slot, and an explicit key is one limit for every copy. Copies of one
run don’t wait for each other’s concurrency; use max_parallel for that.
Other steps can’t read a matrix step’s outputs or inputs, since there is
one per copy.See: StepMatrix
ConcurrencyConfig
Concurrency configuration for controlling step execution across variantsSee: ConcurrencyConfig
enum
required
Allowed values:
terraform:planGitSource
EnvironmentVariablesValue
Environment variables available during execution — plain strings or AWS secret referencesSee: EnvironmentVariablesValue(templateable)
enum
required
IaC tool implementation used to execute this stepAllowed values:
opentofu | terraform(templateable)string
required
Terraform/OpenTofu version to execute(minLen:1; templateable)
string | null
Stack ID used for terraform resource/execution tracking
Automatically set when triggered from a stack(templateable)
string[] | null
Integrations whose credentials are injected into the Terraform/OpenTofu process.
Each entry is an integration ID or givenId; at most one per provider.(templateable)
TerraformVariablesValue
Terraform input variables written to a .auto.tfvars.json file during plan creation — supports all HCL types and AWS secret referencesSee: TerraformVariablesValue(templateable)
string[] | null
Terraform tfvars file paths passed to terraform plan as -var-file flags, in the configured order.
Paths are relative to this step’s Terraform source base path.(templateable)
string | null
S3 directory URI to upload the plan file to — overrides the default convention-based path (e.g. s3://bucket/terraform-plans/run-123)(templateable)
enum
default:"apply"
Type of plan to generate — “apply” (default) creates/updates resources, “destroy” tears them downAllowed values:
apply | destroy(templateable)StepInfrastructure
required
Execution environment configuration — required because Terraform plan runs on provisioned computeSee: StepInfrastructure(templateable)