WP Manifestindependent plugin directory
manifest / ai / ai-provider-for-command-code

AI Provider for Command Code

WordPress AI Provider Plugin for Command Code

by Bestony · github.com/bestony/ai-provider-for-command-code · website

★ 0stars
1forks

Install

No release zip yet. The repository archive installs, but the folder name will carry the branch suffix and updates will not flow:

wp plugin install https://github.com/bestony/ai-provider-for-command-code/archive/refs/heads/main.zip
Readme github.com/bestony/ai-provider-for-command-code ↗

English · 简体中文

Command Code as a provider for the WordPress AI Client: Claude, GPT, Gemini and leading open models through one API key.

What it does

  • Fetches the model list live from GET /provider/v1/models, so newly added models show up on their own.
  • Calls Claude models on the Anthropic Messages endpoint and everything else on the OpenAI-compatible chat completions endpoint, matching the supported_endpoints field the API reports per model.
  • Supports tool calling, structured output (JSON schema), chat history, stop sequences and multi-candidate generation wherever the model supports them.

Requirements

  • WordPress 6.9 or newer, with the PHP AI Client SDK (bundled in WordPress 7.0, or provided by the AI plugin)
  • PHP 7.4 or newer
  • A Command Code plan with API access — the free Go plan has none; Provider, GOAT, Pro, Max and Team do

Install

Download the zip from Releases and upload it through Plugins → Add New → Upload Plugin, or copy the plugin folder to wp-content/plugins/ai-provider-for-command-code/. Activate it, then open Settings → Connectors, open the Command Code card and paste your Provider API key.

Configuration

The API key is read in order of precedence from:

  1. the COMMANDCODE_API_KEY environment variable
  2. the COMMANDCODE_API_KEY PHP constant
  3. the connectors_ai_commandcode_api_key option (Settings → Connectors)

Optional settings are environment variables or PHP constants:

Setting Default Purpose
COMMANDCODE_DEFAULT_MODEL claude-sonnet-5 Model ID to prefer in pickers and feature filters
COMMANDCODE_ZDR off 1 sends x-cmd-zdr: 1, routing only through zero-data-retention upstreams; a model with no ZDR upstream fails with 422 instead of falling back
COMMANDCODE_BASE_URL https://api.commandcode.ai/provider/v1 Provider API base URL
COMMANDCODE_MODEL_INPUT_MODALITIES text only Comma-separated input modalities, e.g. text,image for vision features
COMMANDCODE_REQUEST_TIMEOUT 120 Request timeout, seconds
COMMANDCODE_CONNECT_TIMEOUT 10 Connection timeout, seconds
COMMANDCODE_STRUCTURED_OUTPUT json_schema json_schema, json_object, or none — use json_object if a JSON feature fails with 400 ... "param":"response_format"

To change which model the AI plugin picks for its features, use the standard filter — this plugin already puts its preferred model first:

add_filter( 'wpai_preferred_text_models', function ( $models ) {
    array_unshift( $models, array( 'commandcode', 'gpt-5.6-luna' ) );
    return $models;
} );

Data and privacy

Prompts you send to the AI, plus whatever the calling plugin or theme adds to them (system instructions, conversation history, tool definitions, a JSON schema, attached files), are sent to Command Code at api.commandcode.ai. Your API key is stored on your own site and is only ever sent to that host. Nothing is sent until your site actually asks the AI Client for a generation.

Development

php scripts/selfcheck.php                                    # logic checks, no WordPress needed
php scripts/selfcheck.php --sdk=/path/to/wordpress-php-ai-client/src   # plus live-SDK checks
scripts/bump-version.sh patch                                # bump patch|minor|major, then add a changelog entry

Release: push a tag matching the plugin version (git tag v1.0.4 && git push origin v1.0.4). The release workflow verifies the tag against Version:/Stable tag:, builds the plugin zip and publishes it with a changelog taken from readme.txt.

License

GPL-2.0-or-later. See LICENSE.

The full WordPress plugin readme, including the changelog, lives in readme.txt.