The stored Command Code API key is no longer read by the plugin; the AI Client reports whether a credential is configured.<\/p>","1.0.3":"
Adds the External services documentation and the GPL license file.<\/p>","1.0.2":"
Plugin Check cleanup: escaped exception messages, autoloader direct-access guard, five readme tags.<\/p>","1.0.1":"
Fix Adds Command Code<\/a> as a provider for the WordPress AI Client.\nCommand Code routes Claude, GPT, Gemini and the top open models through a single API with\nOpenAI-compatible and Anthropic-compatible endpoints, so one API key covers every model the\nAI plugin can use.<\/p>\n\n The API key is read, in order of precedence, from:<\/p>\n\n All optional settings are environment variables or PHP constants:<\/p>\n\n To influence which models the AI plugin picks for its features, use the standard filters in your own\nplugin or theme \u2014 this plugin already puts its preferred model first:<\/p>\n\n This plugin connects to the Command Code Provider API, an external service operated by Command Code.\nIt is required so the WordPress AI Client can route requests to models from your site. Command Code is\na paid service: a plan with API access is needed (the free Go plan has none), and requests are billed\nto your Command Code account.<\/p>\n\n The plugin contacts the following endpoints under No request is made until something on your site asks the AI Client for a generation, or the AI Client\nrefreshes its model list. Your API key is stored on your own site and is only ever sent to\n api.commandcode.ai.<\/p>\n\n Optional: setting This service is provided by Command Code:<\/p>\n\n A Command Code plan with API access is required. The Go plan has no API access; Provider, GOAT, Pro,\nMax and Team plans do.<\/p>\n\n\n No. It requires the PHP AI Client SDK, which is provided by WordPress 7.0 or by the AI plugin. The\nprovider stays silent when the SDK is missing.<\/p><\/dd>\n Only the API key, and only if you cannot set an environment variable or constant. The provider adds\nno settings page of its own: everything it needs is on Settings \u2192 Connectors.<\/p><\/dd>\n Command Code's models endpoint does not report which models accept images, so the plugin declares\nimage input from a maintained list in the source. If your model supports images but is not on that\nlist, set GPT-5 and later reasoning families reject non-default sampling parameters, so the plugin does not\nadvertise them for those models. That is upstream behaviour, not a plugin limitation.<\/p><\/dd>\n That feature asks for a JSON schema response. Set In the WordPress options table on your own site, under the AI Client's connector option\n( Only what you send to the AI: your prompts, and whatever the calling plugin or theme adds to them\n(system instructions, conversation history, tool definitions, a JSON schema, or attached files). See\nExternal services below for the exact endpoints.<\/p><\/dd>\n No. Everything this plugin needs lives on Settings \u2192 Connectors, and the optional behaviour is\ncontrolled by environment variables or constants.<\/p><\/dd>\n\n<\/dl>\n\n\n400 invalid_request_error: response_format<\/code> on JSON schema features such as Editorial Notes.<\/p>"},"ratings":[],"assets_icons":{"icon.svg":{"filename":"icon.svg","revision":3706444,"resolution":false,"location":"assets","locale":false}},"assets_banners":{"banner-1544x500.png":{"filename":"banner-1544x500.png","revision":3706444,"resolution":"1544x500","location":"assets","locale":"","width":1544,"height":500},"banner-772x250.png":{"filename":"banner-772x250.png","revision":3706444,"resolution":"772x250","location":"assets","locale":"","width":772,"height":250}},"assets_blueprints":{},"all_blocks":[],"tagged_versions":["1.0.6"],"block_files":[],"assets_screenshots":{"screenshot-1.png":{"filename":"screenshot-1.png","revision":3706444,"resolution":"1","location":"assets","locale":"","width":2072,"height":1262}},"screenshots":{"1":"The Command Code connector card on Settings \u2192 Connectors, with the API key field and the\nmodels the provider reports."}},"plugin_section":[],"plugin_tags":[2353,229563,281925,25308,216551],"plugin_category":[],"plugin_contributors":[145631],"plugin_business_model":[],"class_list":["post-371597","plugin","type-plugin","status-publish","hentry","plugin_tags-ai","plugin_tags-claude","plugin_tags-command-code","plugin_tags-connector","plugin_tags-gpt","plugin_contributors-bestony","plugin_committers-bestony"],"banners":{"banner":"https:\/\/ps.w.org\/ai-provider-for-command-code\/assets\/banner-772x250.png?rev=3706444","banner_2x":"https:\/\/ps.w.org\/ai-provider-for-command-code\/assets\/banner-1544x500.png?rev=3706444","banner_rtl":false,"banner_2x_rtl":false},"icons":{"svg":"https:\/\/ps.w.org\/ai-provider-for-command-code\/assets\/icon.svg?rev=3706444","icon":"https:\/\/ps.w.org\/ai-provider-for-command-code\/assets\/icon.svg?rev=3706444","icon_2x":false,"generated":false},"screenshots":[{"src":"https:\/\/ps.w.org\/ai-provider-for-command-code\/assets\/screenshot-1.png?rev=3706444","caption":"The Command Code connector card on Settings \u2192 Connectors, with the API key field and the\nmodels the provider reports."}],"raw_content":"\n\n
GET \/provider\/v1\/models<\/code>, including newly added models.<\/li>\nsupported_endpoints<\/code> field the API\nreports for each model.<\/li>\nConfiguration<\/h3>\n\n
\n
COMMANDCODE_API_KEY<\/code> environment variable<\/li>\nCOMMANDCODE_API_KEY<\/code> PHP constant<\/li>\nconnectors_ai_commandcode_api_key<\/code> option (Settings \u2192 Connectors)<\/li>\n<\/ol>\n\n\n
COMMANDCODE_DEFAULT_MODEL<\/code> \u2014 model ID to prefer in pickers and feature filters.\nDefault: claude-sonnet-5<\/code>.<\/li>\nCOMMANDCODE_ZDR<\/code> \u2014 set to 1<\/code> to send x-cmd-zdr: 1<\/code> and route only through zero-data-retention\nupstreams. Requests fail with a 422 when the chosen model has no ZDR-capable upstream instead of\nsilently falling back. Default: off.<\/li>\nCOMMANDCODE_BASE_URL<\/code> \u2014 Provider API base URL.\nDefault: https:\/\/api.commandcode.ai\/provider\/v1<\/code>.<\/li>\nCOMMANDCODE_MODEL_INPUT_MODALITIES<\/code> \u2014 comma-separated list of modalities your models accept.\nInclude image<\/code> (e.g. text,image<\/code>) for deployments that require vision features. Default: text only.<\/li>\nCOMMANDCODE_REQUEST_TIMEOUT<\/code> \u2014 request timeout in seconds. Default: 120<\/code>.<\/li>\nCOMMANDCODE_CONNECT_TIMEOUT<\/code> \u2014 connection timeout in seconds. Default: 10<\/code>.<\/li>\nCOMMANDCODE_STRUCTURED_OUTPUT<\/code> \u2014 how JSON response requests are shaped. One of:\n\n\n
json_schema<\/code> (default) \u2014 sends the JSON schema. Use this by default.<\/li>\njson_object<\/code> \u2014 asks only for valid JSON, without the schema. Widely supported fallback.<\/li>\nnone<\/code> \u2014 sends no response_format<\/code> at all. Use this when a model rejects every form of it.\nIf a feature that needs JSON (e.g. Editorial Notes) fails with\n400 ... \"param\":\"response_format\", try json_object<\/code>.<\/li>\n<\/ul><\/li>\n<\/ul>\n\nadd_filter( 'wpai_preferred_text_models', function ( $models ) {\n array_unshift( $models, array( 'commandcode', 'gpt-5.6-luna' ) );\n return $models;\n} );<h3>External services<\/h3>\n<\/code><\/pre>\n\nhttps:\/\/api.commandcode.ai\/provider\/v1<\/code>:<\/p>\n\n\n
GET \/models<\/code> \u2014 called when the AI Client refreshes its list of available models, and when it checks\nwhether your credentials work. No user content is sent; only your API key, so Command Code can\nreturn the models available to your account.<\/li>\nPOST \/chat\/completions<\/code> \u2014 called whenever any plugin or theme on your site uses the WordPress AI\nClient to generate text with a non-Claude model. The request carries your API key and the prompt:\nmessages, system instruction, tool definitions, and any other parameters the calling code supplied\n(conversation history, a JSON schema for structured output, or files attached to the prompt).<\/li>\nPOST \/messages<\/code> \u2014 the same as above, for Claude models, which Command Code serves only on this\nendpoint.<\/li>\n<\/ul>\n\nCOMMANDCODE_ZDR=1<\/code> adds an x-cmd-zdr: 1<\/code> header, asking Command Code to route the\nrequest only through zero-data-retention upstreams. Requests fail with a 422 when a chosen model has\nno such upstream, rather than falling back silently.<\/p>\n\n\n
\n
\/wp-content\/plugins\/ai-provider-for-command-code\/<\/code>.<\/li>\n\n
Does this plugin work without the PHP AI Client?<\/h3><\/dt>\n
Do I need to configure anything in the database?<\/h3><\/dt>\n
Which models can be used with vision features?<\/h3><\/dt>\n
COMMANDCODE_MODEL_INPUT_MODALITIES=text,image<\/code> to declare it.<\/p><\/dd>\nWhy is `temperature` missing on GPT-5 models?<\/h3><\/dt>\n
Editorial Notes fails with `400 ... response_format`<\/h3><\/dt>\n
COMMANDCODE_STRUCTURED_OUTPUT=json_object<\/code> to ask\nfor plain JSON instead, or none<\/code> to send no response_format<\/code> at all \u2014 the prompt still asks for JSON.<\/p><\/dd>\nWhere is my API key stored?<\/h3><\/dt>\n
connectors_ai_commandcode_api_key<\/code>), or in an environment variable or PHP constant if you set one.\nIt is never transmitted anywhere except to the Command Code API when fulfilling a request.<\/p><\/dd>\nWhat data leaves my site?<\/h3><\/dt>\n
Is there a settings page?<\/h3><\/dt>\n
1.0.6<\/h4>\n\n
\n
connectors_ai_commandcode_api_key<\/code> option directly. Whether a Command Code\ncredential is configured is now asked of the AI Client, so the plugin never handles the key the\nuser saved in Settings \u2192 Connectors. Behaviour is unchanged: the COMMANDCODE_API_KEY<\/code> constant and\nenvironment variable still work.<\/li>\n<\/ul>\n\n1.0.3<\/h4>\n\n
\n
1.0.2<\/h4>\n\n
\n
1.0.1<\/h4>\n\n
\n
400 invalid_request_error: response_format<\/code> on features that request a JSON schema, such as\nEditorial Notes. The JSON schema is now sent in the wrapper the API expects.<\/li>\n<\/ul>\n\n1.0.0<\/h4>\n\n
\n