AI response: after provider call

Plugin ID: ai_eca_interceptor:ai_response

Available since: 2.0.0

Provided tokens

Token	Description
[user]	The current user. Alias: current_user
[htmx]	Information about the current HTMX request.
[htmx:is_request]	Whether the current request was sent by HTMX ("1" or "0").
[htmx:boosted]	Whether the current request was boosted by HTMX ("1" or "0").
[htmx:history_restore]	Whether the current request is for HTMX history restoration ("1" or "0").
[htmx:target]	The identifier of the element the response is targeted at.
[htmx:trigger]	The identifier of the element that triggered the request. On Drupal 11 this is the element id, on Drupal 12 and later it is the CSS selector that htmx 4 sends. Prefer "trigger_name", which means the same thing on both.
[htmx:trigger_name]	The name attribute of the element that triggered the request, or empty if it has none. This has the same meaning on all supported Drupal versions.
[htmx:source]	The CSS selector of the element that triggered the request, for example button[name="first_item"]. Empty on Drupal 11, which does not send this information.
[htmx:request_type]	Whether the request targets a full page or a fragment: "full" or "partial". Empty on Drupal 11, which does not send this information.
[htmx:current_url]	The URL of the page the request was sent from.
[event]	The AI event.
[event:machine_name]	Machine name of the triggered ECA event (e.g. ai.pre_generate_response).
[event:request_thread_id]	Unique id of the AI request thread.
[event:request_parent_id]	Parent request id if this request was initiated from another AI request.
[event:provider_id]	The AI provider id that will execute (or did execute) the call.
[event:model_id]	The AI model id.
[event:operation_type]	Operation type: chat, embeddings, moderation, speech_to_text, text_to_speech, etc.
[event:tags]	Tags attached to the request (array).
[event:tag_count]	Number of tags attached to the request.
[event:configuration]	Provider configuration array (temperature, max_tokens, …).
[event:debug_data]	Debug data attached to the request.
[event:metadata]	All metadata on the event, including reroute directives.
[event:is_chat]	1 when the operation type is "chat", 0 otherwise.
[event:input]	The raw input DTO flattened to an array (input->toArray()).
[event:input_class]	Fully qualified class name of the input DTO.
[event:input_text]	Best-effort textual representation of the input (chat prompts joined, summarization/translation/moderation text, etc.).
[event:chat_tools]	Chat only. The declared tools rendered to an array.
[event:chat_tool_names]	Chat only. The names of the declared tool functions (array).
[event:chat_tool_count]	Chat only. Number of declared tool functions.
[event:chat_has_tools]	Chat only. 1 when the request declares any tools, 0 otherwise.
[event:chat_has_schema]	Chat only. 1 when the request declares a structured JSON schema, 0 otherwise.
[event:chat_streamed]	Chat only. 1 when the request asked for streamed output, 0 otherwise.
[event:output]	Response only. The normalized output converted to an array.
[event:output_class]	Response only. Fully qualified class name of the output DTO.
[event:output_normalized]	Response only. The normalized output (array form where possible).
[event:output_raw]	Response only. The provider's raw output payload.
[event:output_metadata]	Response only. Metadata returned by the provider.
[event:output_text]	Response only. Best-effort textual representation of the output.
[event:output_is_streamed]	Response only, chat. 1 when the output is a streamed iterator, 0 otherwise.
[event:token_usage]	Response only, chat. Token usage DTO broken out to fields.
[event:token_usage:input]	Input tokens consumed.
[event:token_usage:output]	Output tokens produced.
[event:token_usage:total]	Total tokens used.
[event:token_usage:reasoning]	Reasoning tokens used (if reported).
[event:token_usage:cached]	Cached tokens reused (if reported).
[event:rate_limits]	Response only, chat. Provider rate limit DTO fields.
[event:rate_limits:max_requests]	Rate limit max requests window.
[event:rate_limits:max_tokens]	Rate limit max tokens window.
[event:rate_limits:remaining_requests]	Remaining requests in the current window.
[event:rate_limits:remaining_tokens]	Remaining tokens in the current window.
[event:rate_limits:reset_requests]	Seconds until the request window resets.
[event:rate_limits:reset_tokens]	Seconds until the token window resets.
[event:moderation_flagged]	Response only, moderation. 1 when the response is flagged, 0 otherwise.
[event:moderation_message]	Response only, moderation. The moderation response message.
[event]	The event.
[event:machine_name]	The machine name of the ECA event.
[session_user]	The user account that dispatched the event, regardless if ECA is processing models under a different account. This is only available if ECA is configured to always run under a specific account.

Fields

Operation type filter — config key: operation_type

Limit to a specific AI operation type (e.g. "chat", "embeddings"). Leave empty for any.

Type: textfield · Required: no · Default: empty string

Provider id filter — config key: provider_id

Limit to a specific provider id. Leave empty for any.

Type: textfield · Required: no · Default: empty string

Model id filter — config key: model_id

Limit to a specific model id. Leave empty for any.

Type: textfield · Required: no · Default: empty string