agenta 0.57.0__py3-none-any.whl → 0.65.0__py3-none-any.whl

agenta/sdk/workflows/builtin.py

@@ -0,0 +1,600 @@
+from typing import Optional, Union, Dict
+
+from agenta.sdk.models.workflows import Reference
+from agenta.sdk.decorators.running import workflow, Workflow, application, evaluator
+from agenta.sdk.workflows.handlers import SinglePromptConfig
+
+
+def echo(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+) -> Workflow:
+    return workflow(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="echo",
+    )()
+
+
+def auto_exact_match(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    correct_answer_key: Optional[str] = "correct_answer",
+) -> Workflow:
+    parameters = dict(
+        correct_answer_key=correct_answer_key,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_exact_match",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_regex_test(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    regex_pattern: str,
+    #
+    regex_should_match: Optional[bool] = True,
+    case_sensitive: Optional[bool] = True,
+) -> Workflow:
+    parameters = dict(
+        regex_pattern=regex_pattern,
+        regex_should_match=regex_should_match,
+        case_sensitive=case_sensitive,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_regex_test",
+        #
+        parameters=parameters,
+    )()
+
+
+def field_match_test(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    json_field: str,
+    #
+    correct_answer_key: Optional[str] = "correct_answer",
+) -> Workflow:
+    parameters = dict(
+        json_field=json_field,
+        correct_answer_key=correct_answer_key,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="field_match_test",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_webhook_test(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    webhook_url: str,
+    #
+    correct_answer_key: Optional[str] = "correct_answer",
+) -> Workflow:
+    parameters = dict(
+        webhook_url=webhook_url,
+        correct_answer_key=correct_answer_key,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_webhook_test",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_custom_code_run(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    code: str,
+    #
+    correct_answer_key: Optional[str] = "correct_answer",
+    threshold: Optional[float] = 0.5,
+) -> Workflow:
+    parameters = dict(
+        code=code,
+        correct_answer_key=correct_answer_key,
+        threshold=threshold,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_custom_code_run",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_ai_critique(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    prompt_template: list[dict[str, str]],
+    #
+    correct_answer_key: Optional[str] = "correct_answer",
+    model: Optional[str] = "gpt-3.5-turbo",
+) -> Workflow:
+    parameters = dict(
+        prompt_template=prompt_template,
+        correct_answer_key=correct_answer_key,
+        model=model,
+        version=3,
+        template_format="curly",
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_ai_critique",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_starts_with(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    prefix: str,
+    #
+    case_sensitive: Optional[bool] = True,
+) -> Workflow:
+    parameters = dict(
+        prefix=prefix,
+        case_sensitive=case_sensitive,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_starts_with",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_ends_with(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    suffix: str,
+    #
+    case_sensitive: Optional[bool] = True,
+) -> Workflow:
+    parameters = dict(
+        suffix=suffix,
+        case_sensitive=case_sensitive,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_ends_with",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_contains(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    substring: str,
+    #
+    case_sensitive: Optional[bool] = True,
+) -> Workflow:
+    parameters = dict(
+        substring=substring,
+        case_sensitive=case_sensitive,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_contains",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_contains_any(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    substrings: list[str],
+    #
+    case_sensitive: Optional[bool] = True,
+) -> Workflow:
+    parameters = dict(
+        substrings=substrings,
+        case_sensitive=case_sensitive,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_contains_any",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_contains_all(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    substrings: list[str],
+    #
+    case_sensitive: Optional[bool] = True,
+) -> Workflow:
+    parameters = dict(
+        substrings=substrings,
+        case_sensitive=case_sensitive,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_contains_all",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_contains_json(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+) -> Workflow:
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_contains_json",
+    )()
+
+
+def auto_json_diff(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    correct_answer_key: Optional[str] = "correct_answer",
+    threshold: Optional[float] = 0.5,
+    predict_keys: Optional[bool] = False,
+    case_insensitive_keys: Optional[bool] = False,
+    compare_schema_only: Optional[bool] = False,
+) -> Workflow:
+    parameters = dict(
+        correct_answer_key=correct_answer_key,
+        threshold=threshold,
+        predict_keys=predict_keys,
+        case_insensitive_keys=case_insensitive_keys,
+        compare_schema_only=compare_schema_only,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_json_diff",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_levenshtein_distance(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    correct_answer_key: Optional[str] = "correct_answer",
+    case_sensitive: Optional[bool] = True,
+    threshold: Optional[float] = 0.5,
+) -> Workflow:
+    parameters = dict(
+        correct_answer_key=correct_answer_key,
+        case_sensitive=case_sensitive,
+        threshold=threshold,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_levenshtein_distance",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_similarity_match(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    correct_answer_key: Optional[str] = "correct_answer",
+    case_sensitive: Optional[bool] = True,
+    threshold: Optional[float] = 0.5,
+) -> Workflow:
+    parameters = dict(
+        correct_answer_key=correct_answer_key,
+        case_sensitive=case_sensitive,
+        threshold=threshold,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_similarity_match",
+        #
+        parameters=parameters,
+    )()
+
+
+def auto_semantic_similarity(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    correct_answer_key: Optional[str] = "correct_answer",
+    threshold: Optional[float] = 0.5,
+    embedding_model: Optional[str] = "text-embedding-3-small",
+) -> Workflow:
+    parameters = dict(
+        correct_answer_key=correct_answer_key,
+        threshold=threshold,
+        embedding_model=embedding_model,
+    )
+
+    return evaluator(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="auto_semantic_similarity",
+        #
+        parameters=parameters,
+    )()
+
+
+def completion(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    config: SinglePromptConfig,
+) -> Workflow:
+    parameters = config.model_dump(
+        mode="json",
+        exclude_none=True,
+    )
+
+    return application(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="completion",
+        #
+        parameters=parameters,
+    )()
+
+
+def chat(
+    *,
+    slug: Optional[str] = None,
+    #
+    name: Optional[str] = None,
+    description: Optional[str] = None,
+    #
+    variant_slug: Optional[str] = None,
+    #
+    config: SinglePromptConfig,
+) -> Workflow:
+    parameters = config.model_dump(
+        mode="json",
+        exclude_none=True,
+    )
+
+    return application(
+        slug=slug,
+        #
+        name=name,
+        description=description,
+        #
+        variant_slug=variant_slug,
+        #
+        uri="chat",
+        #
+        parameters=parameters,
+    )()

agenta/sdk/workflows/configurations.py

@@ -0,0 +1,22 @@
+from agenta.sdk.models.workflows import WorkflowServiceConfiguration
+
+
+echo_v0_configuration = WorkflowServiceConfiguration()
+auto_exact_match_v0_configuration = WorkflowServiceConfiguration()
+auto_regex_test_v0_configuration = WorkflowServiceConfiguration()
+field_match_test_v0_configuration = WorkflowServiceConfiguration()
+auto_webhook_test_v0_configuration = WorkflowServiceConfiguration()
+auto_custom_code_run_v0_configuration = WorkflowServiceConfiguration()
+auto_ai_critique_v0_configuration = WorkflowServiceConfiguration()
+auto_starts_with_v0_configuration = WorkflowServiceConfiguration()
+auto_ends_with_v0_configuration = WorkflowServiceConfiguration()
+auto_contains_v0_configuration = WorkflowServiceConfiguration()
+auto_contains_any_v0_configuration = WorkflowServiceConfiguration()
+auto_contains_all_v0_configuration = WorkflowServiceConfiguration()
+auto_contains_json_v0_configuration = WorkflowServiceConfiguration()
+auto_json_diff_v0_configuration = WorkflowServiceConfiguration()
+auto_levenshtein_distance_v0_configuration = WorkflowServiceConfiguration()
+auto_similarity_match_v0_configuration = WorkflowServiceConfiguration()
+auto_semantic_similarity_v0_configuration = WorkflowServiceConfiguration()
+completion_v0_configuration = WorkflowServiceConfiguration()
+chat_v0_configuration = WorkflowServiceConfiguration()