jeans/developers.home-assistant
jeans/developers.home-assistant
1
0
Fork 0
mirror of https://github.com/home-assistant/developers.home-assistant.git synced 2025-07-14 12:56:30 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add attachments

Browse Source
This commit is contained in:
Paulus Schoutsen 2025-07-06 17:50:05 +00:00
parent 7dc11465b4
commit 4fa0a48d88
1 changed files with 6 additions and 54 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

60
docs/core/entity/ai-task.md
View File

@ -39,7 +39,11 @@ class MyAITaskEntity(AITaskEntity):
# Use the chat_log to maintain conversation context. A common # Use the chat_log to maintain conversation context. A common
# pattern is to share an implementation between conversation and AI # pattern is to share an implementation between conversation and AI
# task entities to process the chat log. # task entities to process the chat log.
# await self._async_handle_chat_log(chat_log) # await self._async_handle_chat_log(
# chat_log,
# task.structure,
# task.attachments
# )
text = ... text = ...
if not task.structure: if not task.structure:
@ -63,59 +67,7 @@ A `GenDataTask` object contains the following data:
| `task_name` | `str` | Name/identifier for the task | `task_name` | `str` | Name/identifier for the task
| `instructions` | `str` | Natural language instructions for the AI | `instructions` | `str` | Natural language instructions for the AI
| `structure` | `dict` \| `None` | Optional schema for structured output validation | `structure` | `dict` \| `None` | Optional schema for structured output validation
| `attachments` | `list[PlayMediaWithId]` | List of attachments to include in the task.
## Services
### Generate Data
The `ai_task.generate_data` service allows users to request AI-generated content or structured data.
#### Parameters
| Name | Type | Required | Description
| ---- | ---- | -------- | -----------
| `task_name` | `str` | Yes | Name/identifier for the task
| `instructions` | `str` | Yes | Natural language instructions for the AI
| `entity_id` | `str` | No | Specific AI task entity to use. Falls back to preferred entity if not specified.
| `structure` | `dict` | No | Schema for structured output validation
#### Service Response
```python
{
"conversation_id": "unique_conversation_id",
"data": "generated_content_or_structured_data"
}
```
#### Examples
**Basic text generation:**
```yaml
service: ai_task.generate_data
data:
task_name: "weather_summary"
instructions: "Generate a funny weather notification"
entity_id: "ai_task.my_ai_assistant"
```
**Structured data generation:**
```yaml
service: ai_task.generate_data
data:
task_name: "user_profile"
instructions: "Create a fake user profile"
structure:
name:
description: "Full name"
required: true
selector:
text: {}
email:
description: "Email address"
selector:
text: {}
```
## Structured Output Schema ## Structured Output Schema