judgeval 0.16.9__tar.gz → 0.18.0__tar.gz

{judgeval-0.16.9 → judgeval-0.18.0}/.github/workflows/ci.yaml

@@ -51,7 +51,7 @@ jobs:
           cd src
           export JUDGMENT_API_KEY="$JUDGEVAL_GH_JUDGMENT_API_KEY"
           export JUDGMENT_ORG_ID="$JUDGEVAL_GH_JUDGMENT_ORG_ID"
-          uv run pytest tests -n auto
+          uv run pytest tests/tracer/llm -n auto
 
   run-e2e-tests:
     needs: [validate-branch]

{judgeval-0.16.9 → judgeval-0.18.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: judgeval
-Version: 0.16.9
+Version: 0.18.0
 Summary: Judgeval Package
 Project-URL: Homepage, https://github.com/JudgmentLabs/judgeval
 Project-URL: Issues, https://github.com/JudgmentLabs/judgeval/issues
@@ -63,8 +63,7 @@ Judgeval's agent monitoring infra provides a simple harness for integrating GRPO
 await trainer.train(
     agent_function=your_agent_function,  # entry point to your agent
     scorers=[RewardScorer()],  # Custom scorer you define based on task criteria, acts as reward
-    prompts=training_prompts,  # Tasks
-    rft_provider="fireworks"
+    prompts=training_prompts  # Tasks
 )
 ```
 

{judgeval-0.16.9 → judgeval-0.18.0}/README.md

@@ -36,8 +36,7 @@ Judgeval's agent monitoring infra provides a simple harness for integrating GRPO
 await trainer.train(
     agent_function=your_agent_function,  # entry point to your agent
     scorers=[RewardScorer()],  # Custom scorer you define based on task criteria, acts as reward
-    prompts=training_prompts,  # Tasks
-    rft_provider="fireworks"
+    prompts=training_prompts  # Tasks
 )
 ```
 

{judgeval-0.16.9 → judgeval-0.18.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "judgeval"
-version = "0.16.9"
+version = "0.18.0"
 authors = [
     { name = "Andrew Li", email = "andrew@judgmentlabs.ai" },
     { name = "Alex Shan", email = "alex@judgmentlabs.ai" },

{judgeval-0.16.9 → judgeval-0.18.0}/scripts/api_generator.py

@@ -43,6 +43,11 @@ JUDGEVAL_PATHS: List[str] = [
     "/e2e_fetch_trace/",
     "/e2e_fetch_span_score/",
     "/e2e_fetch_trace_scorer_span_score/",
+    "/prompts/insert/",
+    "/prompts/fetch/",
+    "/prompts/tag/",
+    "/prompts/untag/",
+    "/prompts/get_prompt_versions/",
 ]
 
 

{judgeval-0.16.9 → judgeval-0.18.0}/scripts/openapi_transform.py

@@ -41,6 +41,11 @@ JUDGEVAL_PATHS: List[str] = [
     "/projects/resolve/",
     "/e2e_fetch_trace/",
     "/e2e_fetch_span_score/",
+    "/prompts/insert/",
+    "/prompts/fetch/",
+    "/prompts/tag/",
+    "/prompts/untag/",
+    "/prompts/get_prompt_versions/",
 ]
 
 

{judgeval-0.16.9 → judgeval-0.18.0}/src/judgeval/__init__.py

@@ -6,6 +6,7 @@ from judgeval.data.evaluation_run import ExampleEvaluationRun
 
 
 from typing import List, Optional, Union, Sequence
+import ast
 from judgeval.scorers import ExampleAPIScorerConfig
 from judgeval.scorers.example_scorer import ExampleScorer
 from judgeval.data.example import Example
@@ -81,6 +82,7 @@ class JudgmentClient(metaclass=SingletonMeta):
         scorer_file_path: str,
         requirements_file_path: Optional[str] = None,
         unique_name: Optional[str] = None,
+        overwrite: bool = False,
     ) -> bool:
         """
         Upload custom ExampleScorer from files to backend.
@@ -89,6 +91,7 @@ class JudgmentClient(metaclass=SingletonMeta):
             scorer_file_path: Path to Python file containing CustomScorer class
             requirements_file_path: Optional path to requirements.txt
             unique_name: Optional unique identifier (auto-detected from scorer.name if not provided)
+            overwrite: Whether to overwrite existing scorer if it already exists
 
         Returns:
             bool: True if upload successful
@@ -111,6 +114,31 @@ class JudgmentClient(metaclass=SingletonMeta):
         with open(scorer_file_path, "r") as f:
             scorer_code = f.read()
 
+        try:
+            tree = ast.parse(scorer_code, filename=scorer_file_path)
+        except SyntaxError as e:
+            error_msg = f"Invalid Python syntax in {scorer_file_path}: {e}"
+            judgeval_logger.error(error_msg)
+            raise ValueError(error_msg)
+
+        scorer_classes = []
+        for node in ast.walk(tree):
+            if isinstance(node, ast.ClassDef):
+                for base in node.bases:
+                    if (isinstance(base, ast.Name) and base.id == "ExampleScorer") or (
+                        isinstance(base, ast.Attribute) and base.attr == "ExampleScorer"
+                    ):
+                        scorer_classes.append(node.name)
+
+        if len(scorer_classes) > 1:
+            error_msg = f"Multiple ExampleScorer classes found in {scorer_file_path}: {scorer_classes}. Please only upload one scorer class per file."
+            judgeval_logger.error(error_msg)
+            raise ValueError(error_msg)
+        elif len(scorer_classes) == 0:
+            error_msg = f"No ExampleScorer class was found in {scorer_file_path}. Please ensure the file contains a valid scorer class that inherits from ExampleScorer."
+            judgeval_logger.error(error_msg)
+            raise ValueError(error_msg)
+
         # Read requirements (optional)
         requirements_text = ""
         if requirements_file_path and os.path.exists(requirements_file_path):
@@ -127,6 +155,7 @@ class JudgmentClient(metaclass=SingletonMeta):
                     "scorer_name": unique_name,
                     "scorer_code": scorer_code,
                     "requirements_text": requirements_text,
+                    "overwrite": overwrite,
                 }
             )
 

{judgeval-0.16.9 → judgeval-0.18.0}/src/judgeval/api/__init__.py

@@ -189,6 +189,59 @@ class JudgmentSyncClient:
             payload,
         )
 
+    def prompts_insert(self, payload: PromptInsertRequest) -> PromptInsertResponse:
+        return self._request(
+            "POST",
+            url_for("/prompts/insert/"),
+            payload,
+        )
+
+    def prompts_tag(self, payload: PromptTagRequest) -> PromptTagResponse:
+        return self._request(
+            "POST",
+            url_for("/prompts/tag/"),
+            payload,
+        )
+
+    def prompts_untag(self, payload: PromptUntagRequest) -> PromptUntagResponse:
+        return self._request(
+            "POST",
+            url_for("/prompts/untag/"),
+            payload,
+        )
+
+    def prompts_fetch(
+        self,
+        project_id: str,
+        name: str,
+        commit_id: Optional[str] = None,
+        tag: Optional[str] = None,
+    ) -> PromptFetchResponse:
+        query_params = {}
+        query_params["project_id"] = project_id
+        query_params["name"] = name
+        if commit_id is not None:
+            query_params["commit_id"] = commit_id
+        if tag is not None:
+            query_params["tag"] = tag
+        return self._request(
+            "GET",
+            url_for("/prompts/fetch/"),
+            query_params,
+        )
+
+    def prompts_get_prompt_versions(
+        self, project_id: str, name: str
+    ) -> PromptVersionsResponse:
+        query_params = {}
+        query_params["project_id"] = project_id
+        query_params["name"] = name
+        return self._request(
+            "GET",
+            url_for("/prompts/get_prompt_versions/"),
+            query_params,
+        )
+
     def projects_resolve(
         self, payload: ResolveProjectNameRequest
     ) -> ResolveProjectNameResponse:
@@ -381,6 +434,61 @@ class JudgmentAsyncClient:
             payload,
         )
 
+    async def prompts_insert(
+        self, payload: PromptInsertRequest
+    ) -> PromptInsertResponse:
+        return await self._request(
+            "POST",
+            url_for("/prompts/insert/"),
+            payload,
+        )
+
+    async def prompts_tag(self, payload: PromptTagRequest) -> PromptTagResponse:
+        return await self._request(
+            "POST",
+            url_for("/prompts/tag/"),
+            payload,
+        )
+
+    async def prompts_untag(self, payload: PromptUntagRequest) -> PromptUntagResponse:
+        return await self._request(
+            "POST",
+            url_for("/prompts/untag/"),
+            payload,
+        )
+
+    async def prompts_fetch(
+        self,
+        project_id: str,
+        name: str,
+        commit_id: Optional[str] = None,
+        tag: Optional[str] = None,
+    ) -> PromptFetchResponse:
+        query_params = {}
+        query_params["project_id"] = project_id
+        query_params["name"] = name
+        if commit_id is not None:
+            query_params["commit_id"] = commit_id
+        if tag is not None:
+            query_params["tag"] = tag
+        return await self._request(
+            "GET",
+            url_for("/prompts/fetch/"),
+            query_params,
+        )
+
+    async def prompts_get_prompt_versions(
+        self, project_id: str, name: str
+    ) -> PromptVersionsResponse:
+        query_params = {}
+        query_params["project_id"] = project_id
+        query_params["name"] = name
+        return await self._request(
+            "GET",
+            url_for("/prompts/get_prompt_versions/"),
+            query_params,
+        )
+
     async def projects_resolve(
         self, payload: ResolveProjectNameRequest
     ) -> ResolveProjectNameResponse:

{judgeval-0.16.9 → judgeval-0.18.0}/src/judgeval/api/api_types.py

@@ -1,6 +1,6 @@
 # generated by datamodel-codegen:
 #   filename:  .openapi.json
-#   timestamp: 2025-10-15T19:25:00+00:00
+#   timestamp: 2025-10-21T01:37:42+00:00
 
 from __future__ import annotations
 from typing import Any, Dict, List, Literal, Optional, TypedDict, Union
@@ -80,6 +80,7 @@ class CustomScorerUploadPayload(TypedDict):
     scorer_name: str
     scorer_code: str
     requirements_text: str
+    overwrite: NotRequired[bool]
 
 
 class CustomScorerTemplateResponse(TypedDict):
@@ -88,6 +89,40 @@ class CustomScorerTemplateResponse(TypedDict):
     message: str
 
 
+class PromptInsertRequest(TypedDict):
+    project_id: str
+    name: str
+    prompt: str
+    tags: List[str]
+
+
+class PromptInsertResponse(TypedDict):
+    commit_id: str
+    parent_commit_id: NotRequired[Optional[str]]
+    created_at: str
+
+
+class PromptTagRequest(TypedDict):
+    project_id: str
+    name: str
+    commit_id: str
+    tags: List[str]
+
+
+class PromptTagResponse(TypedDict):
+    commit_id: str
+
+
+class PromptUntagRequest(TypedDict):
+    project_id: str
+    name: str
+    tags: List[str]
+
+
+class PromptUntagResponse(TypedDict):
+    commit_ids: List[str]
+
+
 class ResolveProjectNameRequest(TypedDict):
     project_name: str
 
@@ -169,6 +204,18 @@ class PromptScorer(TypedDict):
     is_trace: NotRequired[Optional[bool]]
 
 
+class PromptCommitInfo(TypedDict):
+    name: str
+    prompt: str
+    tags: List[str]
+    commit_id: str
+    parent_commit_id: NotRequired[Optional[str]]
+    created_at: str
+    first_name: str
+    last_name: str
+    user_email: str
+
+
 class ScorerData(TypedDict):
     id: NotRequired[str]
     name: str
@@ -265,6 +312,14 @@ class FetchPromptScorersResponse(TypedDict):
     scorers: List[PromptScorer]
 
 
+class PromptFetchResponse(TypedDict):
+    commit: NotRequired[Optional[PromptCommitInfo]]
+
+
+class PromptVersionsResponse(TypedDict):
+    versions: List[PromptCommitInfo]
+
+
 class ScoringResult(TypedDict):
     success: bool
     scorers_data: List[ScorerData]

{judgeval-0.16.9 → judgeval-0.18.0}/src/judgeval/cli.py

@@ -26,6 +26,12 @@ def upload_scorer(
     unique_name: str = typer.Option(
         None, help="Custom name for the scorer (auto-detected if not provided)"
     ),
+    overwrite: bool = typer.Option(
+        False,
+        "--overwrite",
+        "-o",
+        help="Overwrite existing scorer if it already exists",
+    ),
 ):
     # Validate file paths
     if not Path(scorer_file_path).exists():
@@ -43,6 +49,7 @@ def upload_scorer(
             scorer_file_path=scorer_file_path,
             requirements_file_path=requirements_file_path,
             unique_name=unique_name,
+            overwrite=overwrite,
         )
 
         if not result:

{judgeval-0.16.9 → judgeval-0.18.0}/src/judgeval/data/judgment_types.py

@@ -1,6 +1,6 @@
 # generated by datamodel-codegen:
 #   filename:  .openapi.json
-#   timestamp: 2025-10-15T19:24:59+00:00
+#   timestamp: 2025-10-21T01:37:41+00:00
 
 from __future__ import annotations
 from typing import Annotated, Any, Dict, List, Optional, Union
@@ -87,6 +87,7 @@ class CustomScorerUploadPayload(BaseModel):
     scorer_name: Annotated[str, Field(title="Scorer Name")]
     scorer_code: Annotated[str, Field(title="Scorer Code")]
     requirements_text: Annotated[str, Field(title="Requirements Text")]
+    overwrite: Annotated[Optional[bool], Field(title="Overwrite")] = False
 
 
 class CustomScorerTemplateResponse(BaseModel):
@@ -95,6 +96,40 @@ class CustomScorerTemplateResponse(BaseModel):
     message: Annotated[str, Field(title="Message")]
 
 
+class PromptInsertRequest(BaseModel):
+    project_id: Annotated[str, Field(title="Project Id")]
+    name: Annotated[str, Field(title="Name")]
+    prompt: Annotated[str, Field(title="Prompt")]
+    tags: Annotated[List[str], Field(title="Tags")]
+
+
+class PromptInsertResponse(BaseModel):
+    commit_id: Annotated[str, Field(title="Commit Id")]
+    parent_commit_id: Annotated[Optional[str], Field(title="Parent Commit Id")] = None
+    created_at: Annotated[str, Field(title="Created At")]
+
+
+class PromptTagRequest(BaseModel):
+    project_id: Annotated[str, Field(title="Project Id")]
+    name: Annotated[str, Field(title="Name")]
+    commit_id: Annotated[str, Field(title="Commit Id")]
+    tags: Annotated[List[str], Field(title="Tags")]
+
+
+class PromptTagResponse(BaseModel):
+    commit_id: Annotated[str, Field(title="Commit Id")]
+
+
+class PromptUntagRequest(BaseModel):
+    project_id: Annotated[str, Field(title="Project Id")]
+    name: Annotated[str, Field(title="Name")]
+    tags: Annotated[List[str], Field(title="Tags")]
+
+
+class PromptUntagResponse(BaseModel):
+    commit_ids: Annotated[List[str], Field(title="Commit Ids")]
+
+
 class ResolveProjectNameRequest(BaseModel):
     project_name: Annotated[str, Field(title="Project Name")]
 
@@ -187,6 +222,18 @@ class PromptScorer(BaseModel):
     is_trace: Annotated[Optional[bool], Field(title="Is Trace")] = False
 
 
+class PromptCommitInfo(BaseModel):
+    name: Annotated[str, Field(title="Name")]
+    prompt: Annotated[str, Field(title="Prompt")]
+    tags: Annotated[List[str], Field(title="Tags")]
+    commit_id: Annotated[str, Field(title="Commit Id")]
+    parent_commit_id: Annotated[Optional[str], Field(title="Parent Commit Id")] = None
+    created_at: Annotated[str, Field(title="Created At")]
+    first_name: Annotated[str, Field(title="First Name")]
+    last_name: Annotated[str, Field(title="Last Name")]
+    user_email: Annotated[str, Field(title="User Email")]
+
+
 class ScorerData(BaseModel):
     id: Annotated[Optional[str], Field(title="Id")] = None
     name: Annotated[str, Field(title="Name")]
@@ -299,6 +346,14 @@ class FetchPromptScorersResponse(BaseModel):
     scorers: Annotated[List[PromptScorer], Field(title="Scorers")]
 
 
+class PromptFetchResponse(BaseModel):
+    commit: Optional[PromptCommitInfo] = None
+
+
+class PromptVersionsResponse(BaseModel):
+    versions: Annotated[List[PromptCommitInfo], Field(title="Versions")]
+
+
 class ScoringResult(BaseModel):
     success: Annotated[bool, Field(title="Success")]
     scorers_data: Annotated[List[ScorerData], Field(title="Scorers Data")]