improved parsing validation

lisadunlap · lisadunlap · commit 3d220a616a98 · 2025-12-07T19:23:54.000Z
diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
@@ -27,14 +27,7 @@ jobs:
         python -m pip install --upgrade pip
         pip install -e .
         pip install -r requirements.txt
-        pip install pytest pytest-cov flake8 black mypy
-
-    - name: Lint with flake8
-      run: |
-        # stop the build if there are Python syntax errors or undefined names
-        flake8 stringsight/ --count --select=E9,F63,F7,F82 --show-source --statistics
-        # exit-zero treats all errors as warnings. The GitHub editor is 127 chars wide
-        flake8 stringsight/ --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics
+        pip install pytest pytest-cov black mypy
 
     - name: Check formatting with black
       run: |
diff --git a/README.md b/README.md
@@ -64,7 +64,7 @@ For local development or contributing, you can install from source in editable m
 
 ```bash
 # Clone the repository with submodules
-git clone --recurse-submodules https://github.com/lisabdunlap/stringsight.git
+git clone --recurse-submodules https://github.com/lisadunlap/stringsight.git
 cd stringsight
 
 # If you already cloned without submodules, initialize them:
@@ -101,7 +101,7 @@ For multi-user deployments or to run StringSight with all infrastructure depende
 
 ```bash
 # Clone the repository
-git clone https://github.com/lisabdunlap/stringsight.git
+git clone https://github.com/lisadunlap/stringsight.git
 cd stringsight
 
 # Copy the environment template and add your API key
@@ -661,4 +661,4 @@ python scripts/run_full_pipeline.py \
 - **Examples**: See `examples/` directory
   
 
-Contributing & Help: PRs welcome. Questions or issues? Open an issue on GitHub (https://github.com/lisabdunlap/stringsight/issues)
+Contributing & Help: PRs welcome. Questions or issues? Open an issue on GitHub (https://github.com/lisadunlap/stringsight/issues)
diff --git a/docs/deployment/production.md b/docs/deployment/production.md
@@ -73,7 +73,7 @@ StringSight includes a complete Docker Compose setup with all infrastructure dep
 
 ```bash
 # Clone repository
-git clone https://github.com/lisabdunlap/stringsight.git
+git clone https://github.com/lisadunlap/stringsight.git
 cd stringsight
 
 # Configure environment
diff --git a/docs/getting-started/installation.md b/docs/getting-started/installation.md
@@ -30,7 +30,7 @@ conda activate stringsight
 pip install "stringsight[full]"
 
 # Or, for local development from source:
-# git clone https://github.com/lisabdunlap/stringsight.git
+# git clone https://github.com/lisadunlap/stringsight.git
 # cd stringsight
 # pip install -e ".[full]"
 ```
@@ -112,7 +112,7 @@ For multi-user deployments or to run StringSight with all infrastructure depende
 
 ```bash
 # Clone the repository
-git clone https://github.com/lisabdunlap/stringsight.git
+git clone https://github.com/lisadunlap/stringsight.git
 cd stringsight
 
 # Copy the environment template and add your API key
diff --git a/docs/index.md b/docs/index.md
@@ -135,8 +135,8 @@ See the [Installation Guide](getting-started/installation.md) for detailed setup
 ## Support
 
 - **Documentation**: You're reading it!
-- **Issues**: [GitHub Issues](https://github.com/lisabdunlap/stringsight/issues)
-- **Source Code**: [GitHub Repository](https://github.com/lisabdunlap/stringsight)
+- **Issues**: [GitHub Issues](https://github.com/lisadunlap/stringsight/issues)
+- **Source Code**: [GitHub Repository](https://github.com/lisadunlap/stringsight)
 
 ## License
 
diff --git a/docs/troubleshooting.md b/docs/troubleshooting.md
@@ -163,6 +163,6 @@ clusterer = HDBSCANClusterer(disable_dim_reduction=True)
 
 ## Getting Help
 
-- **GitHub Issues**: [Report bugs](https://github.com/lisabdunlap/stringsight/issues)
+- **GitHub Issues**: [Report bugs](https://github.com/lisadunlap/stringsight/issues)
 - **Documentation**: [Read the docs](index.md)
 - **Logs**: Check console output for error details
diff --git a/mkdocs.yml b/mkdocs.yml
@@ -4,7 +4,7 @@ site_author: StringSight Team
 
 # Repository
 repo_name: StringSight
-repo_url: https://github.com/lisabdunlap/StringSight
+repo_url: https://github.com/lisadunlap/StringSight
 edit_uri: edit/main/docs/
 
 # Development server address
diff --git a/pyproject.toml b/pyproject.toml
@@ -57,7 +57,6 @@ dev = [
     "pytest>=6.0",
     "pytest-cov>=2.0",
     "black>=21.0",
-    "flake8>=3.8",
     "mypy>=0.900",
 ]
 local-embeddings = [
@@ -80,10 +79,10 @@ full = [
 ]
 
 [project.urls]
-Homepage = "https://github.com/lisabdunlap/stringsight"
-Repository = "https://github.com/lisabdunlap/stringsight"
-Documentation = "https://github.com/lisabdunlap/stringsight/blob/main/README.md"
-"Bug Tracker" = "https://github.com/lisabdunlap/stringsight/issues"
+Homepage = "https://github.com/lisadunlap/stringsight"
+Repository = "https://github.com/lisadunlap/stringsight"
+Documentation = "https://github.com/lisadunlap/stringsight/blob/main/README.md"
+"Bug Tracker" = "https://github.com/lisadunlap/stringsight/issues"
 
 [tool.setuptools.packages.find]
 include = ["stringsight*"]
diff --git a/stringsight/api.py b/stringsight/api.py
@@ -43,7 +43,6 @@
 from datetime import datetime, timedelta
 from datetime import datetime, timedelta
 import hashlib
-from stringsight.routers.auth import get_current_user_optional
 
 logger = get_logger(__name__)
 
@@ -359,10 +358,8 @@ def _resolve_df_and_method(
     expose_headers=["*"],  # Expose all headers to frontend
 )
 
-from stringsight.routers.auth import router as auth_router
 from stringsight.routers.jobs import router as jobs_router
 
-app.include_router(auth_router)
 app.include_router(jobs_router)
 
 # Include metrics endpoints (basic file serving)
@@ -561,7 +558,6 @@ class ClusterRunRequest(BaseModel):
 async def cluster_run(
     req: ClusterRunRequest,
     background_tasks: BackgroundTasks,
-    current_user: Optional[User] = Depends(get_current_user_optional)
 ) -> Dict[str, Any]:
     """Run clustering directly on existing properties without re-running extraction.
     
diff --git a/stringsight/auth.py b/stringsight/auth.py
diff --git a/stringsight/models/__init__.py b/stringsight/models/__init__.py
@@ -1,2 +1,2 @@
-from stringsight.models.user import User
 from stringsight.models.job import Job
+from stringsight.models.user import User
diff --git a/stringsight/models/user.py b/stringsight/models/user.py
@@ -2,6 +2,7 @@
 from sqlalchemy import Column, String, Boolean, DateTime
 from sqlalchemy.dialects.postgresql import UUID
 from sqlalchemy.sql import func
+from sqlalchemy.orm import relationship
 from stringsight.database import Base
 
 class User(Base):
@@ -12,3 +13,7 @@ class User(Base):
     hashed_password = Column(String, nullable=False)
     is_active = Column(Boolean, default=True)
     created_at = Column(DateTime(timezone=True), server_default=func.now())
+    
+    # Relationships are defined in the other models using backref or explicitly here
+    # Job model defines: user = relationship("User", backref="jobs")
+
diff --git a/stringsight/postprocess/validator.py b/stringsight/postprocess/validator.py
@@ -123,4 +123,18 @@ def _save_stage_results(self, data: PropertyDataset, valid_properties: List[Prop
     def _is_valid_property(self, prop: Property) -> bool:
         """Check if a property is valid."""
         # Basic validation - property description should exist and not be empty
-        return bool(prop.property_description and prop.property_description.strip()) 
+        if not (prop.property_description and prop.property_description.strip()):
+            return False
+
+        # Validate behavior_type if present
+        if prop.behavior_type:
+            allowed_types = {
+                "Positive",
+                "Negative (critical)",
+                "Negative (non-critical)",
+                "Style"
+            }
+            if prop.behavior_type not in allowed_types:
+                return False
+
+        return True 
diff --git a/stringsight/prompts/extraction/universal.py b/stringsight/prompts/extraction/universal.py
@@ -26,6 +26,25 @@
 
 ### DEFINITIONS & RUBRIC
 
+You will return a list of json objects, where each object represents a single, distinct property found in the model's response. Each json object should have the following fields:
+
+```json
+[
+  {
+    "behavior_type": "Negative (critical)|Negative (non-critical)|Positive|Style",
+    "property_description": "lowercase verb + exact action + trigger + consequence/policy impact (1-3 sentences)",
+    "category": "1-4 word category (e.g., 'Regex Failure', 'Safety Robustness', 'Response to Jailbreaking Attempts')",
+    "evidence": "exact quote one", "exact quote two", "exact quote three",
+    "reason": "1-2 sentence explanation of why this property is notable or important",
+    "contains_errors": "True|False",
+    "unexpected_behavior": "True|False"
+  },
+  ...
+]
+```
+
+Below are in detail the definitions and rules for each field:
+
 {model_naming_rule}1. BEHAVIOR TYPES
 * **Negative (Critical):** Direct causes of task failure, hallucinations, gibberish, or safety violations.
 * **Negative (Non-Critical):** Inefficiencies, formatting slips, or partial errors that were rectified later that do not cause complete failure.
@@ -154,13 +173,13 @@
     
     "json_schema": """[
   {
-    "property_description": "string (following the Property Description Formula in Section 2: [lowercase verb] + [specific trigger/context] + [consequence])",
-    "category": "string (short category, e.g., 'Tool Use', 'Tone', 'Safety')",
-    "reason": "string (Why does this matter to a developer?)",
-    "evidence": "exact quote 1", "exact quote 2", "exact quote 3",
+    "property_description": "lowercase verb + exact action + trigger + consequence/policy impact (1-3 sentences, exactly like the examples in Section 2: [lowercase verb] + [specific trigger/context] + [consequence])",
+    "category": "1-4 word category (e.g., 'Refund Policy Violation', 'Safety Refusal', 'Deception Handling', 'Internal Reasoning Leak', 'Manipulation Resistance')",
+    "reason": "Why this property is notable/important — explain impact only (1-2 sentences)",
+    "evidence": "exact quote one", "exact quote two", "exact quote three",
     "behavior_type": "Negative (critical)|Negative (non-critical)|Positive|Style",
-    "contains_errors": boolean,
-    "unexpected_behavior": boolean
+    "contains_errors": True|False,
+    "unexpected_behavior": True|False
   }
 ]""",
 
diff --git a/stringsight/routers/auth.py b/stringsight/routers/auth.py
diff --git a/stringsight/routers/jobs.py b/stringsight/routers/jobs.py
