fin-infra 0.1.60__tar.gz → 0.1.62__tar.gz

{fin_infra-0.1.60 → fin_infra-0.1.62}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: fin-infra
-Version: 0.1.60
+Version: 0.1.62
 Summary: Financial infrastructure toolkit: banking connections, market data, credit, cashflows, and brokerage integrations
 License: MIT
 Keywords: finance,banking,plaid,brokerage,markets,credit,tax,cashflow,fintech,infra
@@ -19,6 +19,7 @@ Classifier: Programming Language :: Python :: 3 :: Only
 Classifier: Topic :: Office/Business :: Financial
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Classifier: Typing :: Typed
+Requires-Dist: ai-infra (>=0.1.142)
 Requires-Dist: cashews[redis] (>=7.0)
 Requires-Dist: ccxt (>=4.0.0)
 Requires-Dist: httpx (>=0.25.0)

{fin_infra-0.1.60 → fin_infra-0.1.62}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "fin-infra"
-version = "0.1.60"
+version = "0.1.62"
 description = "Financial infrastructure toolkit: banking connections, market data, credit, cashflows, and brokerage integrations"
 authors = ["Ali Khatami <aliikhatami94@gmail.com>"]
 license = "MIT"
@@ -55,6 +55,7 @@ svc-infra = ">=0.1.0"  # Published version for CI/production
 yahooquery = ">=2.3.0"
 ccxt = ">=4.0.0"
 plaid-python = ">=25.0.0"
+ai-infra = ">=0.1.142"
 
 [tool.poetry.group.dev.dependencies]
 pytest = ">=8.0.0"

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/analytics/models.py

@@ -114,7 +114,7 @@ class SpendingInsight(BaseModel):
 class PersonalizedSpendingAdvice(BaseModel):
     """LLM-generated personalized spending advice.
 
-    Uses ai-infra CoreLLM for structured output generation.
+    Uses ai-infra LLM for structured output generation.
     """
 
     model_config = ConfigDict(extra="forbid")

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/analytics/spending.py

@@ -12,7 +12,7 @@ Generic Applicability:
 
 Features:
 - Statistical analysis: Top merchants, category breakdowns, trends, anomalies
-- LLM-powered insights: Personalized recommendations using ai-infra CoreLLM
+- LLM-powered insights: Personalized recommendations using ai-infra LLM
 - Graceful degradation: Falls back to rule-based insights if LLM unavailable
 - Cost-effective: Structured output for predictable token usage (<$0.01/insight)
 
@@ -427,12 +427,12 @@ async def generate_spending_insights(
 ) -> "PersonalizedSpendingAdvice":
     """Generate personalized spending insights using LLM.
 
-    Uses ai-infra CoreLLM for structured output generation with financial context.
+    Uses ai-infra LLM for structured output generation with financial context.
 
     Args:
         spending_insight: Analyzed spending data from analyze_spending()
         user_context: Optional context (income, goals, budget, preferences)
-        llm_provider: Optional CoreLLM instance (defaults to Google Gemini)
+        llm_provider: Optional LLM instance (defaults to Google Gemini)
 
     Returns:
         PersonalizedSpendingAdvice with LLM-generated recommendations
@@ -468,14 +468,14 @@ async def generate_spending_insights(
 
     # Try to import ai-infra LLM (optional dependency)
     try:
-        from ai_infra.llm import CoreLLM
+        from ai_infra.llm import LLM
     except ImportError:
         # Graceful degradation: return rule-based insights
         return _generate_rule_based_insights(spending_insight, user_context)
 
     # Initialize LLM if not provided
     if llm_provider is None:
-        llm_provider = CoreLLM()
+        llm_provider = LLM()
 
     # Build financial context prompt
     prompt = _build_spending_insights_prompt(spending_insight, user_context)

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/budgets/__init__.py

@@ -87,11 +87,11 @@ __all__ = [
 def __getattr__(name: str):
     """Lazy import for budgets module components."""
     if name == "easy_budgets":
-        from fin_infra.budgets.ease import easy_budgets  # type: ignore[attr-defined]
+        from fin_infra.budgets.ease import easy_budgets
 
         return easy_budgets
     elif name == "add_budgets":
-        from fin_infra.budgets.add import add_budgets  # type: ignore[attr-defined]
+        from fin_infra.budgets.add import add_budgets
 
         return add_budgets
     elif name in (
@@ -119,11 +119,11 @@ def __getattr__(name: str):
 
         return BudgetTracker
     elif name == "check_budget_alerts":
-        from fin_infra.budgets.alerts import check_budget_alerts  # type: ignore[attr-defined]
+        from fin_infra.budgets.alerts import check_budget_alerts
 
         return check_budget_alerts
     elif name == "apply_template":
-        from fin_infra.budgets.templates import apply_template  # type: ignore[attr-defined]
+        from fin_infra.budgets.templates import apply_template
 
         return apply_template
 

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/categorization/llm_layer.py

@@ -20,7 +20,7 @@ from pydantic import BaseModel, Field
 
 # ai-infra imports
 try:
-    from ai_infra.llm import CoreLLM as LLM
+    from ai_infra.llm import LLM
     from ai_infra.llm.providers import Providers
 except ImportError:
     raise ImportError("ai-infra not installed. Install with: pip install ai-infra")

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/chat/__init__.py

@@ -10,10 +10,10 @@ NOT tied to net worth specifically - works across ALL fin-infra domains:
 
 Example:
     from fin_infra.conversation import FinancialPlanningConversation
-    from ai_infra.llm import CoreLLM
+    from ai_infra.llm import LLM
     from svc_infra.cache import get_cache
 
-    llm = CoreLLM()
+    llm = LLM()
     cache = get_cache()
     conversation = FinancialPlanningConversation(
         llm=llm,
@@ -101,7 +101,7 @@ def add_financial_conversation(
 
     Integration:
         - Uses user_router (requires authentication)
-        - Powered by ai-infra CoreLLM (multi-provider support)
+        - Powered by ai-infra LLM (multi-provider support)
         - Uses svc-infra cache for conversation history (24h TTL)
         - Cost: ~$0.018/user/month with Google Gemini
 

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/chat/ease.py

@@ -13,8 +13,8 @@ Example:
     conversation = easy_financial_conversation(provider="openai")
 
     # Custom LLM instance
-    from ai_infra.llm import CoreLLM
-    llm = CoreLLM(temperature=0.3)
+    from ai_infra.llm import LLM
+    llm = LLM(temperature=0.3)
     conversation = easy_financial_conversation(llm=llm)
 """
 
@@ -33,12 +33,12 @@ def easy_financial_conversation(
     Easy builder for financial planning conversation.
 
     One-call setup with sensible defaults:
-    - LLM: ai-infra CoreLLM (Google Gemini default)
+    - LLM: ai-infra LLM (Google Gemini default)
     - Cache: svc-infra cache (24h TTL)
     - Provider: Google (cheapest, $0.018/user/month)
 
     Args:
-        llm: Optional ai-infra CoreLLM instance (auto-created if None)
+        llm: Optional ai-infra LLM instance (auto-created if None)
         cache: Optional svc-infra cache instance (auto-created if None)
         provider: LLM provider ("google", "openai", "anthropic")
         model_name: Optional model name override (uses provider defaults)
@@ -62,16 +62,16 @@ def easy_financial_conversation(
         conversation = easy_financial_conversation(provider="openai")
 
         # Custom LLM instance
-        from ai_infra.llm import CoreLLM
-        llm = CoreLLM(temperature=0.3)
+        from ai_infra.llm import LLM
+        llm = LLM(temperature=0.3)
         conversation = easy_financial_conversation(llm=llm)
     """
     # Auto-create LLM if not provided
     if llm is None:
         try:
-            from ai_infra.llm import CoreLLM
+            from ai_infra.llm import LLM
 
-            llm = CoreLLM()
+            llm = LLM()
         except ImportError:
             raise ImportError("ai-infra not installed. Install with: pip install ai-infra")
 

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/chat/planning.py

@@ -10,19 +10,19 @@ Provides conversational interface for financial Q&A:
 - Personalized advice (uses current net worth, goals, historical data)
 - Natural dialogue (flexible responses, not forced JSON structure)
 
-**Design Choice**: Uses `CoreLLM.achat()` for natural conversation (NOT `with_structured_output()`).
+**Design Choice**: Uses `LLM.achat()` for natural conversation (NOT `with_structured_output()`).
 Conversation should be flexible and natural, not rigidly structured. Other modules (insights,
 categorization, goals) correctly use structured output because they need predictable schemas.
 
-Uses ai-infra CoreLLM for natural conversation.
+Uses ai-infra LLM for natural conversation.
 Caches conversation context for 24h (target: $0.018/user/month cost).
 
 Example:
-    from ai_infra.llm import CoreLLM
+    from ai_infra.llm import LLM
     from svc_infra.cache import get_cache
     from fin_infra.conversation import FinancialPlanningConversation
 
-    llm = CoreLLM()
+    llm = LLM()
     cache = get_cache()
     conversation = FinancialPlanningConversation(
         llm=llm,
@@ -196,10 +196,10 @@ class FinancialPlanningConversation:
     Cost: ~$0.0054/conversation (10 turns with context caching)
 
     Example:
-        from ai_infra.llm import CoreLLM
+        from ai_infra.llm import LLM
         from svc_infra.cache import get_cache
 
-        llm = CoreLLM()
+        llm = LLM()
         cache = get_cache()
         conversation = FinancialPlanningConversation(
             llm=llm,
@@ -232,7 +232,7 @@ class FinancialPlanningConversation:
         Initialize conversation manager.
 
         Args:
-            llm: ai-infra CoreLLM instance
+            llm: ai-infra LLM instance
             cache: svc-infra cache instance (for context storage)
             provider: LLM provider ("google", "openai", "anthropic")
             model_name: Model name (default: gemini-2.0-flash-exp)

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/crypto/insights.py

@@ -1,9 +1,9 @@
 """Crypto portfolio insights using ai-infra LLM.
 
 This module generates personalized insights for cryptocurrency holdings using
-ai-infra's CoreLLM for intelligent analysis and recommendations.
+ai-infra's LLM for intelligent analysis and recommendations.
 
-CRITICAL: Uses ai-infra.llm.CoreLLM (NEVER custom LLM clients).
+CRITICAL: Uses ai-infra.llm.LLM (NEVER custom LLM clients).
 """
 
 from __future__ import annotations
@@ -15,7 +15,7 @@ from typing import TYPE_CHECKING
 from pydantic import BaseModel, Field
 
 if TYPE_CHECKING:
-    from ai_infra.llm import CoreLLM
+    from ai_infra.llm import LLM
 
 
 class CryptoInsight(BaseModel):
@@ -56,13 +56,13 @@ class CryptoHolding(BaseModel):
 async def generate_crypto_insights(
     user_id: str,
     holdings: list[CryptoHolding],
-    llm: CoreLLM | None = None,
+    llm: LLM | None = None,
     total_portfolio_value: Decimal | None = None,
 ) -> list[CryptoInsight]:
     """
     Generate personalized crypto insights using ai-infra LLM.
 
-    Uses ai-infra.llm.CoreLLM for intelligent analysis based on:
+    Uses ai-infra.llm.LLM for intelligent analysis based on:
     - Portfolio concentration (diversification recommendations)
     - Performance trends (gains/losses)
     - Risk assessment (volatility, allocation %)
@@ -71,15 +71,15 @@ async def generate_crypto_insights(
     Args:
         user_id: User identifier
         holdings: List of crypto holdings
-        llm: CoreLLM instance (if None, uses default Google Gemini)
+        llm: LLM instance (if None, uses default Google Gemini)
         total_portfolio_value: Total portfolio value including non-crypto assets
 
     Returns:
         List of personalized CryptoInsight objects
 
     Example:
-        >>> from ai_infra.llm import CoreLLM
-        >>> llm = CoreLLM(provider="google_genai", model="gemini-2.0-flash-exp")
+        >>> from ai_infra.llm import LLM
+        >>> llm = LLM(provider="google_genai", model="gemini-2.0-flash-exp")
         >>> holdings = [
         ...     CryptoHolding(
         ...         symbol="BTC",
@@ -206,14 +206,14 @@ async def _generate_llm_insights(
     holdings: list[CryptoHolding],
     total_crypto_value: Decimal,
     total_portfolio_value: Decimal | None,
-    llm: CoreLLM,
+    llm: LLM,
 ) -> list[CryptoInsight]:
     """
-    Generate AI-powered insights using ai-infra CoreLLM.
+    Generate AI-powered insights using ai-infra LLM.
 
     Uses natural language conversation (NO output_schema) for personalized advice.
 
-    CRITICAL: Uses ai-infra.llm.CoreLLM (never custom LLM clients).
+    CRITICAL: Uses ai-infra.llm.LLM (never custom LLM clients).
     """
     insights = []
 
@@ -260,7 +260,7 @@ Provide your insight:"""
         # Use natural language conversation (no output_schema)
         # Note: In tests, achat is mocked with messages= parameter
         # In production, this should use user_msg, provider, model_name parameters
-        response = await llm.achat(  # type: ignore[call-arg]
+        response = await llm.achat(
             messages=[{"role": "user", "content": prompt}],
         )
 

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/documents/analysis.py

@@ -2,7 +2,7 @@
 AI-powered document analysis and insights.
 
 Uses rule-based analysis (simulated AI) for financial documents.
-Production: Use ai-infra CoreLLM for real AI-powered insights.
+Production: Use ai-infra LLM for real AI-powered insights.
 
 Quick Start:
     >>> from fin_infra.documents.analysis import analyze_document
@@ -14,7 +14,7 @@ Quick Start:
     >>> print(result.recommendations)
 
 Production Integration:
-    - Use ai-infra CoreLLM for all LLM calls (never custom clients)
+    - Use ai-infra LLM for all LLM calls (never custom clients)
     - Cache analysis results (24h TTL via svc-infra cache)
     - Track LLM costs per user (ai-infra cost tracking)
     - Add disclaimers for financial advice
@@ -68,7 +68,7 @@ async def analyze_document(
 
     Notes:
         - Current: Rule-based analysis (simulated AI)
-        - Production: Use ai-infra CoreLLM (never custom LLM clients)
+        - Production: Use ai-infra LLM (never custom LLM clients)
         - Production: Check cache before analysis (svc-infra cache, 24h TTL)
         - Production: Track LLM costs (ai-infra cost tracking)
         - Production: Add disclaimer: "Not a substitute for certified financial advisor"
@@ -218,7 +218,7 @@ def _analyze_tax_document(ocr_text: str, metadata: dict, document_id: str) -> "D
 
     Notes:
         - Current: Rule-based analysis (simulated AI)
-        - Production: Use ai-infra CoreLLM with financial tax prompt
+        - Production: Use ai-infra LLM with financial tax prompt
         - Production: Include tax bracket information
         - Production: Suggest W-4 adjustments if applicable
         - Production: Identify potential deductions or credits
@@ -318,7 +318,7 @@ def _analyze_bank_statement(ocr_text: str, metadata: dict, document_id: str) ->
 
     Notes:
         - Current: Rule-based analysis (simulated AI)
-        - Production: Use ai-infra CoreLLM with spending analysis prompt
+        - Production: Use ai-infra LLM with spending analysis prompt
         - Production: Identify unusual transactions or patterns
         - Production: Compare to typical spending (if available)
         - Production: Suggest budget optimizations

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/goals/management.py

@@ -7,14 +7,14 @@ Provides 4 goal types with validation and progress tracking:
 3. Debt-Free Goal: Debt payoff schedule with APR calculations
 4. Wealth Milestone: Project growth to target net worth
 
-Uses ai-infra CoreLLM for validation + local math for calculations.
+Uses ai-infra LLM for validation + local math for calculations.
 Weekly progress check-ins via svc-infra.jobs scheduler ($0.0036/user/month).
 
 Example:
-    from ai_infra.llm import CoreLLM
+    from ai_infra.llm import LLM
     from fin_infra.goals.management import FinancialGoalTracker
 
-    llm = CoreLLM()
+    llm = LLM()
     tracker = FinancialGoalTracker(
         llm=llm,
         provider="google",
@@ -475,9 +475,9 @@ class FinancialGoalTracker:
     Cost: ~$0.0009/validation, ~$0.0009/week for progress ($0.0036/user/month)
 
     Example:
-        from ai_infra.llm import CoreLLM
+        from ai_infra.llm import LLM
 
-        llm = CoreLLM()
+        llm = LLM()
         tracker = FinancialGoalTracker(llm=llm, provider="google")
 
         # Validate goal
@@ -503,7 +503,7 @@ class FinancialGoalTracker:
         Initialize goal tracker.
 
         Args:
-            llm: ai-infra CoreLLM instance
+            llm: ai-infra LLM instance
             provider: LLM provider ("google", "openai", "anthropic")
             model_name: Model name (default: gemini-2.0-flash-exp)
         """

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/investments/models.py

@@ -201,7 +201,7 @@ class Holding(BaseModel):
     unofficial_currency_code: Optional[str] = Field(None, description="For crypto/alt currencies")
     as_of_date: Optional[date] = Field(None, description="Date of pricing data")
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def unrealized_gain_loss(self) -> Optional[Decimal]:
         """Calculate unrealized gain/loss (current value - cost basis)."""
@@ -209,7 +209,7 @@ class Holding(BaseModel):
             return None
         return self.institution_value - self.cost_basis
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def unrealized_gain_loss_percent(self) -> Optional[Decimal]:
         """Calculate unrealized gain/loss percentage."""
@@ -350,7 +350,7 @@ class InvestmentAccount(BaseModel):
     # Holdings
     holdings: List[Holding] = Field(default_factory=list, description="List of holdings in account")
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def total_value(self) -> Decimal:
         """Calculate total account value (sum of holdings + cash)."""
@@ -358,20 +358,20 @@ class InvestmentAccount(BaseModel):
         cash_balance = self.balances.get("current") or Decimal(0)
         return holdings_value + cash_balance
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def total_cost_basis(self) -> Decimal:
         """Calculate total cost basis (sum of cost_basis across holdings)."""
         return sum(h.cost_basis for h in self.holdings if h.cost_basis is not None)
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def total_unrealized_gain_loss(self) -> Decimal:
         """Calculate total unrealized P&L (value - cost_basis)."""
         holdings_value = sum(h.institution_value for h in self.holdings)
         return holdings_value - self.total_cost_basis
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def total_unrealized_gain_loss_percent(self) -> Optional[Decimal]:
         """Calculate total unrealized P&L percentage."""

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/net_worth/ease.py

@@ -368,7 +368,7 @@ def easy_net_worth(
 
     if enable_llm:
         try:
-            from ai_infra.llm import CoreLLM
+            from ai_infra.llm import LLM
         except ImportError:
             raise ImportError(
                 "LLM features require ai-infra package. " "Install with: pip install ai-infra"
@@ -387,8 +387,8 @@ def easy_net_worth(
                 f"Unknown llm_provider: {llm_provider}. " f"Use 'google', 'openai', or 'anthropic'"
             )
 
-        # Create shared CoreLLM instance
-        llm = CoreLLM()
+        # Create shared LLM instance
+        llm = LLM()
 
         # Create LLM components (deferred import to avoid circular dependency)
         # These modules will be created in Section 17 V2 implementation

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/net_worth/insights.py

@@ -7,14 +7,14 @@ Provides 4 types of insights:
 3. Goal Recommendations: Validate financial goals and suggest paths
 4. Asset Allocation Advice: Portfolio rebalancing based on age/risk
 
-Uses ai-infra CoreLLM with structured output (Pydantic schemas).
+Uses ai-infra LLM with structured output (Pydantic schemas).
 Caches insights for 24h (target: 95%+ hit rate, $0.042/user/month cost).
 
 Example:
-    from ai_infra.llm import CoreLLM
+    from ai_infra.llm import LLM
     from fin_infra.net_worth.insights import NetWorthInsightsGenerator
 
-    llm = CoreLLM()
+    llm = LLM()
     generator = NetWorthInsightsGenerator(
         llm=llm,
         provider="google",
@@ -379,7 +379,7 @@ class NetWorthInsightsGenerator:
     """
     Generate LLM-powered financial insights for net worth tracking.
 
-    Uses ai-infra CoreLLM with structured output (Pydantic schemas).
+    Uses ai-infra LLM with structured output (Pydantic schemas).
     Supports 4 insight types:
     1. Wealth trends (analyze net worth changes)
     2. Debt reduction (prioritize payoff by APR)
@@ -389,9 +389,9 @@ class NetWorthInsightsGenerator:
     Cost: ~$0.042/user/month (1 insight/day, 24h cache, Google Gemini)
 
     Example:
-        from ai_infra.llm import CoreLLM
+        from ai_infra.llm import LLM
 
-        llm = CoreLLM()
+        llm = LLM()
         generator = NetWorthInsightsGenerator(
             llm=llm,
             provider="google",
@@ -412,7 +412,7 @@ class NetWorthInsightsGenerator:
         Initialize insights generator.
 
         Args:
-            llm: ai-infra CoreLLM instance
+            llm: ai-infra LLM instance
             provider: LLM provider ("google", "openai", "anthropic")
             model_name: Model name (default: gemini-2.0-flash-exp)
         """

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/net_worth/models.py

@@ -207,7 +207,7 @@ class AssetAllocation(BaseModel):
     vehicles: float = Field(0.0, ge=0, description="Vehicle value")
     other_assets: float = Field(0.0, ge=0, description="Other asset value")
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def total_assets(self) -> float:
         """Sum of all asset categories."""
@@ -220,37 +220,37 @@ class AssetAllocation(BaseModel):
             + self.other_assets
         )
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def cash_percentage(self) -> float:
         """Cash as percentage of total assets."""
         return (self.cash / self.total_assets * 100) if self.total_assets > 0 else 0.0
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def investments_percentage(self) -> float:
         """Investments as percentage of total assets."""
         return (self.investments / self.total_assets * 100) if self.total_assets > 0 else 0.0
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def crypto_percentage(self) -> float:
         """Crypto as percentage of total assets."""
         return (self.crypto / self.total_assets * 100) if self.total_assets > 0 else 0.0
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def real_estate_percentage(self) -> float:
         """Real estate as percentage of total assets."""
         return (self.real_estate / self.total_assets * 100) if self.total_assets > 0 else 0.0
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def vehicles_percentage(self) -> float:
         """Vehicles as percentage of total assets."""
         return (self.vehicles / self.total_assets * 100) if self.total_assets > 0 else 0.0
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def other_percentage(self) -> float:
         """Other assets as percentage of total assets."""
@@ -288,7 +288,7 @@ class LiabilityBreakdown(BaseModel):
     personal_loans: float = Field(0.0, ge=0, description="Personal loan balance")
     lines_of_credit: float = Field(0.0, ge=0, description="Line of credit balance")
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def total_liabilities(self) -> float:
         """Sum of all liability categories."""
@@ -301,7 +301,7 @@ class LiabilityBreakdown(BaseModel):
             + self.lines_of_credit
         )
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def credit_cards_percentage(self) -> float:
         """Credit cards as percentage of total liabilities."""
@@ -311,7 +311,7 @@ class LiabilityBreakdown(BaseModel):
             else 0.0
         )
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def mortgages_percentage(self) -> float:
         """Mortgages as percentage of total liabilities."""
@@ -319,7 +319,7 @@ class LiabilityBreakdown(BaseModel):
             (self.mortgages / self.total_liabilities * 100) if self.total_liabilities > 0 else 0.0
         )
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def auto_loans_percentage(self) -> float:
         """Auto loans as percentage of total liabilities."""
@@ -327,7 +327,7 @@ class LiabilityBreakdown(BaseModel):
             (self.auto_loans / self.total_liabilities * 100) if self.total_liabilities > 0 else 0.0
         )
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def student_loans_percentage(self) -> float:
         """Student loans as percentage of total liabilities."""
@@ -337,7 +337,7 @@ class LiabilityBreakdown(BaseModel):
             else 0.0
         )
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def personal_loans_percentage(self) -> float:
         """Personal loans as percentage of total liabilities."""
@@ -347,7 +347,7 @@ class LiabilityBreakdown(BaseModel):
             else 0.0
         )
 
-    @computed_field  # type: ignore[misc]
+    @computed_field
     @property
     def lines_of_credit_percentage(self) -> float:
         """Lines of credit as percentage of total liabilities."""

{fin_infra-0.1.60 → fin_infra-0.1.62}/src/fin_infra/obs/classifier.py

@@ -37,6 +37,8 @@ Usage:
 
 from __future__ import annotations
 
+from typing import Callable
+
 # Financial capability prefix patterns (extensible)
 FINANCIAL_ROUTE_PREFIXES = (
     "/banking",
@@ -110,9 +112,9 @@ def financial_route_classifier(route_path: str, method: str) -> str:
 
 
 def compose_classifiers(
-    *classifiers: callable,
+    *classifiers: Callable[[str], str],
     default: str = "public",
-) -> callable:
+) -> Callable[[str], str]:
     """
     Compose multiple route classifiers with fallback logic.