---
title: "Returns the system and user prompt templates for the given LLM assistant task in all supported languages"
method: POST
path: "/llm_assistant/create_prompt_templates"
tags: ["llm_assistant"]
---

# Returns the system and user prompt templates for the given LLM assistant task in all supported languages

`POST /llm_assistant/create_prompt_templates`

## Request

- The document declares no server URL.
- Auth: OAuth 2
  - Password
  - Token URL: `/authentication/login`

## Query parameters

- `approach_type` 'LLM_ZERO_SHOT' | 'LLM_FEW_SHOT', required
- `strategy_type` 'TAGGING_DEFAULT' | 'METADATA_DEFAULT' | 'SENTENCE_ANNOTATION_DEFAULT' | 'NER_INLINE_TAGS' | 'CONTEXT_ANCHORED_FUZZY_MATCHING', required

## Request body

- BodyLlmAssistantCreatePromptTemplates
  - `job_params` LLMAssistantJobParameters, required
    - `project_id` integer, required: Project ID associated with the job
    - `task_type` 'TAGGING' | 'METADATA_EXTRACTION' | 'ANNOTATION' | 'SENTENCE_ANNOTATION', required
    - `specific_task_parameters` union, required: Specific parameters for the LLMAssistantJob w.r.t it's type
      - TaggingParams
        - `task_type` 'TAGGING', required: The type of the LLM Assistant task
        - `sdoc_ids` integer[], required: IDs of the source documents to analyse
        - `tag_ids` integer[], required: IDs of the tags to use for the document tagging
      - MetadataExtractionParams
        - `task_type` 'METADATA_EXTRACTION', required: The type of the LLM Assistant task
        - `sdoc_ids` integer[], required: IDs of the source documents to analyse
        - `project_metadata_ids` integer[], required: IDs of the project metadata to use for the metadata extraction
      - AnnotationParams
        - `task_type` 'ANNOTATION', required: The type of the LLM Assistant task
        - `sdoc_ids` integer[], required: IDs of the source documents to analyse
        - `code_ids` integer[], required: IDs of the codes to use for the annotation
        - `delete_existing_annotations` boolean: Delete existing annotations before creating new ones
      - SentenceAnnotationParams
        - `task_type` 'SENTENCE_ANNOTATION', required: The type of the LLM Assistant task
        - `sdoc_ids` integer[], required: IDs of the source documents to analyse
        - `code_ids` integer[], required: IDs of the codes to use for the sentence annotation
        - `delete_existing_annotations` boolean: Delete existing annotations before creating new ones
  - `example_ids` integer[], nullable

## Response `200`

Successful Response

- LLMAssistantPromptTemplates[]
  - `language` string, required: The language of the prompt template
  - `system_prompt` string, required: The system prompt to use for the job
  - `user_prompt` string, required: The user prompt to use for the job

## Other responses

- `422`: Validation Error

## Changes

- **2026-10-19** `90c133703342`: 1 info
  - endpoint added

[Operation history](https://skmtc.dev/uhh-lt/apis/discourse-analysis-tool-suite-api/changes/llm_assistant/create_prompt_templates/post.md)

---

[API](https://skmtc.dev/uhh-lt/apis/discourse-analysis-tool-suite-api.md) - [All operations](https://skmtc.dev/uhh-lt/apis/discourse-analysis-tool-suite-api/llms.txt) - [OpenAPI document](https://skmtc.dev/uhh-lt/apis/discourse-analysis-tool-suite-api/revisions/e498207e29f4?raw)