diff --git a/.kiro/specs/active/024-generic-agent-interface-contrib/tasks.md b/.kiro/specs/active/024-generic-agent-interface-contrib/tasks.md
index 940deccc..daf56c3b 100644
--- a/.kiro/specs/active/024-generic-agent-interface-contrib/tasks.md
+++ b/.kiro/specs/active/024-generic-agent-interface-contrib/tasks.md
@@ -12,7 +12,7 @@ No standardized input/output contract for agents. Hard to swap, compose, or benc
 - [x] 1. Define AgentInput schema: ticker, date, context (market data, news, fundamentals)
 - [x] 2. Define AgentOutput schema: rating (5-tier), confidence, price_targets, thesis, risk_factors
 - [x] 3. Create BaseAgent abstract class with analyze(input) -> output contract
-- [ ] 4. Refactor existing agents (fundamentals, sentiment, news, technical) to implement BaseAgent
+- [x] 4. Refactor existing agents (fundamentals, sentiment, news, technical) to implement BaseAgent
 - [ ] 5. Create AgentRegistry for pluggable agent discovery
 - [ ] 6. Add agent benchmarking: compare outputs across different LLM backends
 - [ ] 7. Document interface for third-party agent contributions
diff --git a/tradingagents/agents/__init__.py b/tradingagents/agents/__init__.py
index c485fb04..0360e6c0 100644
--- a/tradingagents/agents/__init__.py
+++ b/tradingagents/agents/__init__.py
@@ -8,6 +8,12 @@ from .analysts.fundamentals_analyst import create_fundamentals_analyst
 from .analysts.market_analyst import create_market_analyst
 from .analysts.news_analyst import create_news_analyst
 from .analysts.social_media_analyst import create_social_media_analyst
+from .analysts.base_analysts import (
+    FundamentalsAgent,
+    SentimentAgent,
+    NewsAgent,
+    TechnicalAgent,
+)
 
 from .researchers.bear_researcher import create_bear_researcher
 from .researchers.bull_researcher import create_bull_researcher
@@ -23,6 +29,10 @@ from .trader.trader import create_trader
 
 __all__ = [
     "BaseAgent",
+    "FundamentalsAgent",
+    "SentimentAgent",
+    "NewsAgent",
+    "TechnicalAgent",
     "FinancialSituationMemory",
     "AgentState",
     "AgentInput",
diff --git a/tradingagents/agents/analysts/base_analysts.py b/tradingagents/agents/analysts/base_analysts.py
new file mode 100644
index 00000000..df755ff0
--- /dev/null
+++ b/tradingagents/agents/analysts/base_analysts.py
@@ -0,0 +1,109 @@
+"""BaseAgent implementations for the four analyst types.
+
+Each class wraps the existing analyst logic behind the standardized
+``BaseAgent.analyze(AgentInput) -> AgentOutput`` contract while the
+original ``create_*`` factory functions remain unchanged for LangGraph
+node compatibility.
+"""
+
+from __future__ import annotations
+
+from langchain_core.messages import HumanMessage
+
+from tradingagents.agents.base_agent import BaseAgent
+from tradingagents.agents.utils.schemas import AgentInput, AgentOutput
+
+# Shared prompt that asks the LLM to return a JSON matching AgentOutput.
+_STRUCTURED_SUFFIX = (
+    "\n\nAfter your analysis, provide a final JSON object with these exact keys:\n"
+    '- "rating": one of "BUY", "OVERWEIGHT", "HOLD", "UNDERWEIGHT", "SELL"\n'
+    '- "confidence": float 0.0-1.0\n'
+    '- "thesis": one-paragraph summary\n'
+    '- "risk_factors": list of strings\n'
+    "Return ONLY the JSON object, no other text."
+)
+
+
+def _invoke_structured(llm, role_prompt: str, agent_input: AgentInput) -> AgentOutput:
+    """Ask *llm* to produce an ``AgentOutput`` via structured output."""
+    full_prompt = (
+        f"{role_prompt}\n\n"
+        f"Ticker: {agent_input.ticker}\n"
+        f"Date: {agent_input.date}\n"
+    )
+    if agent_input.context:
+        for k, v in agent_input.context.items():
+            full_prompt += f"\n--- {k} ---\n{v}\n"
+
+    full_prompt += _STRUCTURED_SUFFIX
+
+    structured_llm = llm.with_structured_output(AgentOutput)
+    return structured_llm.invoke([HumanMessage(content=full_prompt)])
+
+
+class FundamentalsAgent(BaseAgent):
+    """Standardized fundamentals analyst."""
+
+    name: str = "fundamentals_analyst"
+
+    def __init__(self, llm) -> None:
+        self.llm = llm
+
+    def analyze(self, agent_input: AgentInput) -> AgentOutput:
+        return _invoke_structured(
+            self.llm,
+            "You are a fundamentals analyst. Evaluate the company's financial health "
+            "using balance sheets, cash flow, income statements, and key ratios.",
+            agent_input,
+        )
+
+
+class SentimentAgent(BaseAgent):
+    """Standardized sentiment / social-media analyst."""
+
+    name: str = "sentiment_analyst"
+
+    def __init__(self, llm) -> None:
+        self.llm = llm
+
+    def analyze(self, agent_input: AgentInput) -> AgentOutput:
+        return _invoke_structured(
+            self.llm,
+            "You are a sentiment analyst. Evaluate public sentiment from social media, "
+            "news headlines, and community discussions about the company.",
+            agent_input,
+        )
+
+
+class NewsAgent(BaseAgent):
+    """Standardized news analyst."""
+
+    name: str = "news_analyst"
+
+    def __init__(self, llm) -> None:
+        self.llm = llm
+
+    def analyze(self, agent_input: AgentInput) -> AgentOutput:
+        return _invoke_structured(
+            self.llm,
+            "You are a news analyst. Evaluate recent news, macroeconomic events, "
+            "and geopolitical developments relevant to the company.",
+            agent_input,
+        )
+
+
+class TechnicalAgent(BaseAgent):
+    """Standardized technical / market analyst."""
+
+    name: str = "technical_analyst"
+
+    def __init__(self, llm) -> None:
+        self.llm = llm
+
+    def analyze(self, agent_input: AgentInput) -> AgentOutput:
+        return _invoke_structured(
+            self.llm,
+            "You are a technical analyst. Evaluate price action, volume, moving averages, "
+            "MACD, RSI, Bollinger Bands, and other technical indicators.",
+            agent_input,
+        )