Skip to content

cli.md

Latest commit

 

History

History
138 lines (106 loc) · 5.14 KB

cli.md

File metadata and controls

138 lines (106 loc) · 5.14 KB
Raw

CLI Command Reference

This document provides a detailed reference for all commands and options available in the wpt-gen CLI.

wpt-gen generate

Generate Web Platform Tests for a specific web feature.

Usage:

wpt-gen generate [OPTIONS] WEB_FEATURE_ID

Arguments

Argument Description
WEB_FEATURE_ID Required. The ID of the feature (e.g., grid, popover) as defined in the web-features repository.

Options

General Options

Option Shorthand Description Default
--provider -p Override the default LLM provider (gemini, openai, anthropic). From config
--wpt-dir -w Override the path to the local web-platform-tests repository. From config
--config -c Path to a custom wpt-gen.yml file. ~/.wpt-gen.yml
--output-dir -o Directory where generated tests will be saved. Local WPT repo
--wf-yml-update Update WEB_FEATURES.yml with generated tests.
--tentative Generate test files with the .tentative flag.

Execution Control

Option Description
--resume Resume the workflow from the last successful phase for this feature ID.
--resume-from Resume the workflow explicitly from a specific phase (context_assembly, requirements_extraction, coverage_audit, generation).
--state-dir Directory containing the necessary artifacts to hydrate the requested phase (alias: --tests-dir).
--suggestions-only Stop after generating test suggestions. Skip the actual test file generation.
--brief-suggestions Only generate test titles and descriptions for suggestions, omitting detailed test suggestions.
--yes-tokens Automatically confirm all prompts related to LLM token counts.
--yes-tests Automatically confirm and generate all proposed test suggestions without prompting.
--yes-cache Automatically use the cache if it exists without prompting.
--no-cache Automatically ignore and overwrite the cache if it exists without prompting.
--max-parallel-requests Limit the number of concurrent asynchronous LLM requests.

Model Configuration

Option Description
--use-lightweight Force the use of the provider's "lightweight" model for all phases.
--use-reasoning Force the use of the provider's "reasoning" model for all phases.
--show-responses, -s Print every raw LLM response to the console (useful for debugging).
--include-thoughts Stream the underlying ADK model thoughts to stdout.
--save-traces Save LLM interaction traces to the .wptgen/traces/ directory.
--temperature Global temperature setting for all LLM requests (e.g., 0.01). Overrides phase-specific defaults.
--max-retries Maximum number of retries for failed LLM calls.
--timeout Timeout for individual LLM requests in seconds.

Feature & Requirement Overrides

Option Shorthand Description
--spec-urls -u Comma-separated list of W3C Spec URLs to use, bypassing automatic fetching.
--description -d Provide a manual description/summary of the feature to the agent.
--detailed-requirements Use an iterative process to extract highly granular requirements (slower).
--draft Enable fetching metadata from the draft features directory.
--include-mdn-docs Include MDN documentation in requirements extraction.
--single-prompt-requirements Use a single-prompt requirements extraction process (legacy).

Local Test Runner

Option Description
--run-on-browser Browser to use for the local WPT test runner (chrome, firefox, safari).
--run-on-channel Release channel to use for the local WPT test runner (canary, nightly, stable, dev).

wpt-gen clear-cache

Clears the local cache of scraped specifications and LLM responses.

Usage:

wpt-gen clear-cache [OPTIONS]

Options

Option Shorthand Description
--config -c Path to a custom wpt-gen.yml file.

wpt-gen version

Print the current version of wpt-gen.

Usage:

wpt-gen version

wpt-gen config

Manage WPT-Gen configuration without manually editing the YAML file. If run without subcommands, it displays the currently active configuration.

Usage:

wpt-gen config [COMMAND]

Commands

Command Description
show Display the currently active, fully resolved configuration.
set Update an individual configuration setting using dot-notation.

wpt-gen config set

Update an individual configuration setting. Modifies the local or global wpt-gen.yml file.

Usage:

wpt-gen config set <KEY> <VALUE> [OPTIONS]

Examples:

wpt-gen config set default_provider openai
wpt-gen config set providers.gemini.default_model gemini-3.7-flash
wpt-gen config set show_responses true

Options

Option Shorthand Description
--config -c Path to a custom wpt-gen.yml file.