Evaluation and Output

Pkl evaluation turns a module into a structured value. For language learning, use the official pkl CLI as the baseline.

Evaluate a Module

pkl eval config.pkl

When no format is set, the default renderer chooses PCF. This is useful while learning because the output keeps Pkl's object/listing/mapping shape visible.

name = "api"port = 8080tags = new Listing {  "pkl"  "configuration"}

Evaluate an Expression

Use -x when you only need one value from a module:

pkl eval -x name config.pklpkl eval -x port config.pkl

Render for a Consumer

Use -f to choose an output format:

pkl eval -f json config.pklpkl eval -f yaml config.pkl

Use -o when the output should be written to a file:

pkl eval -f json -o config.json config.pkl

Validation

pkl eval config.pkl

Evaluation must parse the module, resolve imports, enforce type annotations, and run constraints before output is produced. If a constraint fails, fix the source instead of treating rendering as a separate step.

Output Formats

Pkl's CLI supports multiple renderers through -f. This documentation treats PCF, JSON, YAML, properties, plist, and related formats as language-facing output contracts.

Diagnostics

Diagnostics surface from the stage that failed: syntax, import resolution, typing, constraints, evaluation, or rendering. Keep the smallest failing module nearby, then compare the behavior with Coverage Status when the local toolchain differs from the official CLI.