refactor(prompts): Unify prompt storage in src/prompts/ (Priority 5) (#127)

Priority 5: Unified prompt storage in src/prompts/

✅ All checks passed | 0 CodeRabbit comments

src/agents/magentic_agents.py

@@ -11,6 +11,10 @@ from src.agents.tools import (
 from src.clients.base import BaseChatClient
 from src.clients.factory import get_chat_client
 from src.config.domain import ResearchDomain, get_domain_config
+from src.prompts.hypothesis import get_system_prompt as get_hypothesis_prompt
+from src.prompts.judge import get_system_prompt as get_judge_prompt
+from src.prompts.report import get_system_prompt as get_report_prompt
+from src.prompts.search import get_system_prompt as get_search_prompt
 
 
 def create_search_agent(
@@ -34,19 +38,7 @@ def create_search_agent(
     return ChatAgent(
         name="SearchAgent",
         description=config.search_agent_description,
-        instructions=f"""You are a biomedical search specialist. When asked to find evidence:
-
-1. Analyze the request to determine what to search for
-2. Extract key search terms (drug names, disease names, mechanisms)
-3. Use the appropriate search tools:
-   - search_pubmed for peer-reviewed papers
-   - search_clinical_trials for clinical studies
-   - search_preprints for cutting-edge findings
-4. Summarize what you found and highlight key evidence
-
-Be thorough - search multiple databases when appropriate.
-Focus on finding: mechanisms of action, clinical evidence, and specific findings
-related to {config.name}.""",
+        instructions=get_search_prompt(domain),
         chat_client=client,
         tools=[search_pubmed, search_clinical_trials, search_preprints],
         temperature=1.0,  # Explicitly set for reasoning model compatibility (o1/o3)
@@ -69,43 +61,11 @@ def create_judge_agent(
         ChatAgent configured for evidence assessment
     """
     client = chat_client or get_chat_client(api_key=api_key)
-    config = get_domain_config(domain)
 
     return ChatAgent(
         name="JudgeAgent",
         description="Evaluates evidence quality and determines if sufficient for synthesis",
-        instructions=f"""{config.judge_system_prompt}
-
-When asked to evaluate:
-
-1. Review all evidence presented in the conversation
-2. Score on two dimensions (0-10 each):
-   - Mechanism Score: How well is the biological mechanism explained?
-   - Clinical Score: How strong is the clinical/preclinical evidence?
-3. Determine if evidence is SUFFICIENT for a final report:
-   - Sufficient: Clear mechanism + supporting clinical data
-   - Insufficient: Gaps in mechanism OR weak clinical evidence
-4. If insufficient, suggest specific search queries to fill gaps
-
-## CRITICAL OUTPUT FORMAT
-To ensure the workflow terminates when appropriate, you MUST follow these rules:
-
-IF evidence is SUFFICIENT (confidence >= 70%):
-   Start your response with a line like:
-   "✅ SUFFICIENT EVIDENCE (confidence: 72%). STOP SEARCHING. Delegate to ReportAgent NOW."
-   Use your actual numeric confidence instead of 72.
-   Then explain why.
-
-IF evidence is INSUFFICIENT:
-   Start with "❌ INSUFFICIENT: <Reason>."
-   Then provide scores and next queries.
-
-Be rigorous but fair. Look for:
-- Molecular targets and pathways
-- Animal model studies
-- Human clinical trials
-- Safety data
-- Drug-drug interactions""",
+        instructions=get_judge_prompt(domain),
         chat_client=client,
         temperature=1.0,  # Explicitly set for reasoning model compatibility
     )
@@ -132,23 +92,7 @@ def create_hypothesis_agent(
     return ChatAgent(
         name="HypothesisAgent",
         description=config.hypothesis_agent_description,
-        instructions=f"""{config.hypothesis_system_prompt}
-
-Based on evidence:
-
-1. Identify the key molecular targets involved
-2. Map the biological pathways affected
-3. Generate testable hypotheses in this format:
-
-   DRUG -> TARGET -> PATHWAY -> THERAPEUTIC EFFECT
-
-   Example:
-   Testosterone -> Androgen receptor -> Dopamine modulation -> Enhanced libido
-
-4. Explain the rationale for each hypothesis
-5. Suggest what additional evidence would support or refute it
-
-Focus on mechanistic plausibility and existing evidence.""",
+        instructions=get_hypothesis_prompt(domain),
         chat_client=client,
         temperature=1.0,  # Explicitly set for reasoning model compatibility
     )
@@ -170,48 +114,11 @@ def create_report_agent(
         ChatAgent configured for report generation
     """
     client = chat_client or get_chat_client(api_key=api_key)
-    config = get_domain_config(domain)
 
     return ChatAgent(
         name="ReportAgent",
         description="Synthesizes research findings into structured reports",
-        instructions=f"""{config.report_system_prompt}
-
-When asked to synthesize:
-
-Generate a structured report with these sections:
-
-## Executive Summary
-Brief overview of findings and recommendation
-
-## Methodology
-Databases searched, queries used, evidence reviewed
-
-## Key Findings
-### Mechanism of Action
-- Molecular targets
-- Biological pathways
-- Proposed mechanism
-
-### Clinical Evidence
-- Preclinical studies
-- Clinical trials
-- Safety profile
-
-## Candidates
-List specific candidates with potential
-
-## Limitations
-Gaps in evidence, conflicting data, caveats
-
-## Conclusion
-Final recommendation with confidence level
-
-## References
-Use the 'get_bibliography' tool to fetch the complete list of citations.
-Format them as a numbered list.
-
-Be comprehensive but concise. Cite evidence for all claims.""",
+        instructions=get_report_prompt(domain),
         chat_client=client,
         tools=[get_bibliography],
         temperature=1.0,  # Explicitly set for reasoning model compatibility

src/config/domain.py

@@ -46,16 +46,6 @@ class DomainConfig(BaseModel):
     report_title: str
     report_focus: str
 
-    # Judge prompts
-    judge_system_prompt: str
-    judge_scoring_prompt: str
-
-    # Hypothesis prompts
-    hypothesis_system_prompt: str
-
-    # Report writer prompts
-    report_system_prompt: str
-
     # Search context
     search_description: str
     search_example_query: str
@@ -74,21 +64,6 @@ SEXUAL_HEALTH_CONFIG = DomainConfig(
     description="Sexual health and wellness research specialist",
     report_title="## Sexual Health Analysis",
     report_focus="sexual health and wellness interventions",
-    judge_system_prompt="""You are an expert sexual health research judge.
-Your role is to evaluate evidence for sexual health interventions, assess
-efficacy and safety data, and determine clinical applicability.""",
-    judge_scoring_prompt="""Score this evidence for sexual health relevance.
-Provide ONLY scores and extracted data.""",
-    hypothesis_system_prompt=(
-        """You are a biomedical research scientist specializing in sexual health.
-Your role is to generate evidence-based hypotheses for sexual health interventions,
-identifying mechanisms of action and potential therapeutic applications."""
-    ),
-    report_system_prompt=(
-        """You are a scientific writer specializing in sexual health research reports.
-Your role is to synthesize evidence into clear recommendations for sexual health
-interventions with proper safety considerations."""
-    ),
     search_description="Searches biomedical literature for sexual health evidence",
     search_example_query="testosterone therapy female libido",
     search_agent_description="Searches PubMed for sexual health evidence",

src/prompts/hypothesis.py

@@ -13,33 +13,36 @@ if TYPE_CHECKING:
 def get_system_prompt(domain: ResearchDomain | str | None = None) -> str:
     """Get the system prompt for the hypothesis agent."""
     config = get_domain_config(domain)
-    return f"""{config.hypothesis_system_prompt}
 
-Your role is to generate mechanistic hypotheses based on evidence.
+    return f"""You are a biomedical research scientist specializing in {config.name}.
+Your role is to generate evidence-based hypotheses for interventions,
+identifying mechanisms of action and potential therapeutic applications.
 
-A good hypothesis:
-1. Proposes a MECHANISM: Drug -> Target -> Pathway -> Effect
-2. Is TESTABLE: Can be supported or refuted by literature search
-3. Is SPECIFIC: Names actual molecular targets and pathways
-4. Generates SEARCH QUERIES: Helps find more evidence
+Based on evidence:
 
-Example hypothesis format:
-- Drug: Testosterone
-- Target: Androgen Receptor
-- Pathway: Dopaminergic signaling modulation
-- Effect: Enhanced libido in HSDD
-- Confidence: 0.7
-- Search suggestions: ["testosterone libido mechanism", "sildenafil efficacy women"]
+1. Identify the key molecular targets involved
+2. Map the biological pathways affected
+3. Generate testable hypotheses in this format:
 
-Be specific. Use actual gene/protein names when possible."""
+   DRUG -> TARGET -> PATHWAY -> THERAPEUTIC EFFECT
 
+   Example:
+   Testosterone -> Androgen receptor -> Dopamine modulation -> Enhanced libido
 
-# Keep SYSTEM_PROMPT for backwards compatibility
+4. Explain the rationale for each hypothesis
+5. Suggest what additional evidence would support or refute it
+
+Focus on mechanistic plausibility and existing evidence."""
+
+
+# Keep SYSTEM_PROMPT for backwards compatibility (used by PydanticAI agents)
 SYSTEM_PROMPT = get_system_prompt()
 
 
 async def format_hypothesis_prompt(
-    query: str, evidence: list["Evidence"], embeddings: "EmbeddingServiceProtocol | None" = None
+    query: str,
+    evidence: list["Evidence"],
+    embeddings: "EmbeddingServiceProtocol | None" = None,
 ) -> str:
     """Format prompt for hypothesis generation.
 

src/prompts/judge.py

@@ -5,73 +5,49 @@ from src.utils.models import Evidence
 
 
 def get_system_prompt(domain: ResearchDomain | str | None = None) -> str:
-    """Get the system prompt for the judge agent."""
+    """Get the system prompt for the judge agent (Magentic/Advanced Mode)."""
     config = get_domain_config(domain)
-    return f"""{config.judge_system_prompt}
-
-Your task is to SCORE evidence from biomedical literature. You do NOT decide whether to
-continue searching or synthesize - that decision is made by the orchestration system
-based on your scores.
-
-## Your Role: Scoring Only
-
-You provide objective scores. The system decides next steps based on explicit thresholds.
-This separation prevents bias in the decision-making process.
-
-## Scoring Criteria
-
-1. **Mechanism Score (0-10)**: How well does the evidence explain the biological mechanism?
-   - 0-3: No clear mechanism, speculative
-   - 4-6: Some mechanistic insight, but gaps exist
-   - 7-10: Clear, well-supported mechanism of action
-
-2. **Clinical Evidence Score (0-10)**: Strength of clinical/preclinical support?
-   - 0-3: No clinical data, only theoretical
-   - 4-6: Preclinical or early clinical data
-   - 7-10: Strong clinical evidence (trials, meta-analyses)
-
-3. **Drug Candidates**: List SPECIFIC drug names mentioned in the evidence
-   - Only include drugs explicitly mentioned
-   - Do NOT hallucinate or infer drug names
-   - Include drug class if specific names aren't available (e.g., "SSRI antidepressants")
-
-4. **Key Findings**: Extract 3-5 key findings from the evidence
-   - Focus on findings relevant to the research question
-   - Include mechanism insights and clinical outcomes
-
-5. **Confidence (0.0-1.0)**: Your confidence in the scores
-   - Based on evidence quality and relevance
-   - Lower if evidence is tangential or low-quality
-
-## Output Format
-
-Return valid JSON with these fields:
-- details.mechanism_score (int 0-10)
-- details.mechanism_reasoning (string)
-- details.clinical_evidence_score (int 0-10)
-- details.clinical_reasoning (string)
-- details.drug_candidates (list of strings)
-- details.key_findings (list of strings)
-- sufficient (boolean) - TRUE if scores suggest enough evidence
-- confidence (float 0-1)
-- recommendation ("continue" or "synthesize") - Your suggestion (system may override)
-- next_search_queries (list) - If continuing, suggest FOCUSED queries
-- reasoning (string)
-
-## CRITICAL: Search Query Rules
-
-When suggesting next_search_queries:
-- STAY FOCUSED on the original research question
-- Do NOT drift to tangential topics
-- If question is about "female libido", do NOT suggest "bone health" or "muscle mass"
-- Refine existing terms, don't explore random medical associations
-"""
+
+    return f"""You are an expert research judge specializing in {config.name}.
+Your role is to evaluate evidence for interventions, assess efficacy and safety data,
+and determine clinical applicability.
+
+When asked to evaluate:
+
+1. Review all evidence presented in the conversation
+2. Score on two dimensions (0-10 each):
+   - Mechanism Score: How well is the biological mechanism explained?
+   - Clinical Score: How strong is the clinical/preclinical evidence?
+3. Determine if evidence is SUFFICIENT for a final report:
+   - Sufficient: Clear mechanism + supporting clinical data
+   - Insufficient: Gaps in mechanism OR weak clinical evidence
+4. If insufficient, suggest specific search queries to fill gaps
+
+## CRITICAL OUTPUT FORMAT
+To ensure the workflow terminates when appropriate, you MUST follow these rules:
+
+IF evidence is SUFFICIENT (confidence >= 70%):
+   Start your response with a line like:
+   "✅ SUFFICIENT EVIDENCE (confidence: 72%). STOP SEARCHING. Delegate to ReportAgent NOW."
+   Use your actual numeric confidence instead of 72.
+   Then explain why.
+
+IF evidence is INSUFFICIENT:
+   Start with "❌ INSUFFICIENT: <Reason>."
+   Then provide scores and next queries.
+
+Be rigorous but fair. Look for:
+- Molecular targets and pathways
+- Animal model studies
+- Human clinical trials
+- Safety data
+- Drug-drug interactions"""
 
 
 def get_scoring_prompt(domain: ResearchDomain | str | None = None) -> str:
     """Get the scoring instructions for the judge."""
-    config = get_domain_config(domain)
-    return config.judge_scoring_prompt
+    return """Score this evidence for relevance.
+Provide ONLY scores and extracted data."""
 
 
 # Keep SYSTEM_PROMPT for backwards compatibility
@@ -118,9 +94,6 @@ def format_user_prompt(
 ) -> str:
     """
     Format user prompt with selected evidence and iteration context.
-
-    NOTE: Evidence should be pre-selected using select_evidence_for_judge().
-    This function assumes evidence is already capped.
     """
     # Use explicit None check - 0 is a valid count (empty evidence)
     total_count = total_evidence_count if total_evidence_count is not None else len(evidence)
@@ -140,7 +113,6 @@ def format_user_prompt(
 
     evidence_text = "\n\n".join([format_single_evidence(i, e) for i, e in enumerate(evidence)])
 
-    # Lost-in-the-middle mitigation: put critical context at START and END
     return f"""## Research Question (IMPORTANT - stay focused on this)
 {question}
 
@@ -156,22 +128,12 @@ def format_user_prompt(
 ## Your Task
 
 {scoring_prompt}
-DO NOT decide "synthesize" vs "continue" - that decision is made by the system.
-
-## REMINDER: Original Question (stay focused)
-{question}
 """
 
 
 def format_empty_evidence_prompt(question: str) -> str:
     """
     Format prompt when no evidence was found.
-
-    Args:
-        question: The user's research question
-
-    Returns:
-        Formatted prompt string
     """
     return f"""## Research Question
 {question}

src/prompts/report.py

@@ -7,73 +7,52 @@ from src.utils.text_utils import select_diverse_evidence, truncate_at_sentence
 
 if TYPE_CHECKING:
     from src.services.embedding_protocol import EmbeddingServiceProtocol
-    from src.utils.models import Evidence, MechanismHypothesis
+    from src.utils.models import Evidence, HypothesisAssessment
 
 
 def get_system_prompt(domain: ResearchDomain | str | None = None) -> str:
     """Get the system prompt for the report agent."""
     config = get_domain_config(domain)
-    return f"""{config.report_system_prompt}
-
-Your role is to synthesize evidence and hypotheses into a clear, structured report.
-
-A good report:
-1. Has a clear EXECUTIVE SUMMARY (one paragraph, key takeaways)
-2. States the RESEARCH QUESTION clearly
-3. Describes METHODOLOGY (what was searched, how)
-4. Evaluates HYPOTHESES with evidence counts
-5. Separates MECHANISTIC and CLINICAL findings
-6. Lists specific DRUG CANDIDATES
-7. Acknowledges LIMITATIONS honestly
-8. Provides a balanced CONCLUSION
-9. Includes properly formatted REFERENCES
-
-Write in scientific but accessible language. Be specific about evidence strength.
-
-─────────────────────────────────────────────────────────────────────────────
-🚨 CRITICAL: REQUIRED JSON STRUCTURE 🚨
-─────────────────────────────────────────────────────────────────────────────
-
-The `hypotheses_tested` field MUST be a LIST of objects, each with these fields:
-- "hypothesis": the hypothesis text
-- "supported": count of supporting evidence (integer)
-- "contradicted": count of contradicting evidence (integer)
-
-Example:
-  hypotheses_tested: [
-    {{"hypothesis": "Testosterone -> AR -> enhanced libido",
-      "supported": 3, "contradicted": 1}},
-    {{"hypothesis": "Sildenafil inhibits PDE5 pathway",
-      "supported": 5, "contradicted": 0}}
-  ]
-
-The `references` field MUST be a LIST of objects, each with these fields:
-- "title": paper title (string)
-- "authors": author names (string)
-- "source": "pubmed" or "web" (string)
-- "url": the EXACT URL from evidence (string)
-
-Example:
-  references: [
-    {{"title": "Testosterone and Libido", "authors": "Smith",
-      "source": "pubmed", "url": "https://pubmed.ncbi.nlm.nih.gov/123/"}}
-  ]
-
-─────────────────────────────────────────────────────────────────────────────
-🚨 CRITICAL CITATION REQUIREMENTS 🚨
-─────────────────────────────────────────────────────────────────────────────
-
-You MUST follow these rules for the References section:
-
-1. You may ONLY cite papers that appear in the Evidence section above
-2. Every reference URL must EXACTLY match a provided evidence URL
-3. Do NOT invent, fabricate, or hallucinate any references
-4. Do NOT modify paper titles, authors, dates, or URLs
-5. If unsure about a citation, OMIT it rather than guess
-6. Copy URLs exactly as provided - do not create similar-looking URLs
-
-VIOLATION OF THESE RULES PRODUCES DANGEROUS MISINFORMATION.
-─────────────────────────────────────────────────────────────────────────────"""
+
+    return f"""You are a scientific writer specializing in {config.name}.
+Your role is to synthesize evidence into clear recommendations for interventions
+with proper safety considerations.
+
+When asked to synthesize:
+
+Generate a structured report with these sections:
+
+## Executive Summary
+Brief overview of findings and recommendation
+
+## Methodology
+Databases searched, queries used, evidence reviewed
+
+## Key Findings
+### Mechanism of Action
+- Molecular targets
+- Biological pathways
+- Proposed mechanism
+
+### Clinical Evidence
+- Preclinical studies
+- Clinical trials
+- Safety profile
+
+## Candidates
+List specific candidates with potential
+
+## Limitations
+Gaps in evidence, conflicting data, caveats
+
+## Conclusion
+Final recommendation with confidence level
+
+## References
+Use the 'get_bibliography' tool to fetch the complete list of citations.
+Format them as a numbered list.
+
+Be comprehensive but concise. Cite evidence for all claims."""
 
 
 # Keep SYSTEM_PROMPT for backwards compatibility
@@ -83,67 +62,61 @@ SYSTEM_PROMPT = get_system_prompt()
 async def format_report_prompt(
     query: str,
     evidence: list["Evidence"],
-    hypotheses: list["MechanismHypothesis"],
-    assessment: dict[str, Any],
+    hypotheses: list["HypothesisAssessment"] | list[Any],
+    assessment: Any,
     metadata: dict[str, Any],
     embeddings: "EmbeddingServiceProtocol | None" = None,
 ) -> str:
     """Format prompt for report generation.
 
-    Includes full evidence details for accurate citation.
+    Args:
+        query: Research query
+        evidence: Collected evidence
+        hypotheses: Generated hypotheses
+        assessment: Judge assessment details
+        metadata: Search metadata
+        embeddings: Optional embedding service for diverse selection
     """
-    # Select diverse evidence (not arbitrary truncation)
-    selected = await select_diverse_evidence(evidence, n=20, query=query, embeddings=embeddings)
-
-    # Include FULL citation details for each evidence item
-    # This helps the LLM create accurate references
-    evidence_lines = []
-    for e in selected:
-        authors = ", ".join(e.citation.authors or ["Unknown"])
-        evidence_lines.append(
-            f"- **Title**: {e.citation.title}\n"
-            f"  **URL**: {e.citation.url}\n"
-            f"  **Authors**: {authors}\n"
-            f"  **Date**: {e.citation.date or 'n.d.'}\n"
-            f"  **Source**: {e.citation.source}\n"
-            f"  **Content**: {truncate_at_sentence(e.content, 200)}\n"
-        )
-    evidence_summary = "\n".join(evidence_lines)
-
+    # Select diverse evidence (max 15 for report)
+    selected = await select_diverse_evidence(evidence, n=15, query=query, embeddings=embeddings)
+
+    evidence_text = "\n".join(
+        [
+            f"- **{e.citation.title}** ({e.citation.source}): "
+            f"{truncate_at_sentence(e.content, 400)}"
+            for e in selected
+        ]
+    )
+
+    # Format hypotheses if available
+    hypotheses_text = "No specific hypotheses generated."
     if hypotheses:
-        hypotheses_lines = []
+        # Handle both Pydantic models and dicts/objects
+        h_list = []
         for h in hypotheses:
-            hypotheses_lines.append(
-                f"- {h.drug} -> {h.target} -> {h.pathway} -> {h.effect} "
-                f"(Confidence: {h.confidence:.0%})"
-            )
-        hypotheses_summary = "\n".join(hypotheses_lines)
-    else:
-        hypotheses_summary = "No hypotheses generated yet."
-
-    sources = ", ".join(metadata.get("sources", []))
-
-    return f"""Generate a structured research report for the following query.
-
-## Original Query
-{query}
-
-## Evidence Collected ({len(selected)} papers, selected for diversity)
-
-{evidence_summary}
-
-## Hypotheses Generated
-{hypotheses_summary}
-
-## Assessment Scores
-- Mechanism Score: {assessment.get("mechanism_score", "N/A")}/10
-- Clinical Evidence Score: {assessment.get("clinical_score", "N/A")}/10
-- Overall Confidence: {assessment.get("confidence", 0):.0%}
-
-## Metadata
-- Sources Searched: {sources}
-- Search Iterations: {metadata.get("iterations", 0)}
-
-Generate a complete ResearchReport with all sections filled in.
-
-REMINDER: Only cite papers from the Evidence section above. Copy URLs exactly."""
+            if hasattr(h, "hypotheses"):
+                for item in h.hypotheses:
+                    h_list.append(f"- {item.drug} -> {item.target} -> {item.effect}")
+            elif isinstance(h, dict):
+                h_list.append(str(h))
+            else:
+                h_list.append(str(h))
+        if h_list:
+            hypotheses_text = "\n".join(h_list)
+
+    return f"""Generate a comprehensive research report for: "{query}""
+
+## Context
+- **Sources Searched**: {", ".join(metadata.get("sources", []))}
+- **Iterations**: {metadata.get("iterations", 0)}
+
+## Evidence ({len(selected)} key papers)
+{evidence_text}
+
+## Generated Hypotheses
+{hypotheses_text}
+
+## Task
+Synthesize this information into a structured report following the Executive Summary format.
+Focus on clinical applicability and safety.
+Use specific citations from the evidence list."""

src/prompts/search.py

@@ -0,0 +1,22 @@
+"""Prompts for Search Agent."""
+
+from src.config.domain import ResearchDomain, get_domain_config
+
+
+def get_system_prompt(domain: ResearchDomain | str | None = None) -> str:
+    """Get the system prompt for the search agent."""
+    config = get_domain_config(domain)
+
+    return f"""You are a biomedical search specialist. When asked to find evidence:
+
+1. Analyze the request to determine what to search for
+2. Extract key search terms (drug names, disease names, mechanisms)
+3. Use the appropriate search tools:
+   - search_pubmed for peer-reviewed papers
+   - search_clinical_trials for clinical studies
+   - search_preprints for cutting-edge findings
+4. Summarize what you found and highlight key evidence
+
+Be thorough - search multiple databases when appropriate.
+Focus on finding: mechanisms of action, clinical evidence, and specific findings
+related to {config.name}."""

tests/unit/agents/test_magentic_agents_domain.py

@@ -29,7 +29,7 @@ class TestMagenticAgentsDomain:
 
         # Verify domain-specific judge system prompt is passed through
         call_kwargs = mock_agent_cls.call_args.kwargs
-        assert SEXUAL_HEALTH_CONFIG.judge_system_prompt in call_kwargs["instructions"]
+        assert SEXUAL_HEALTH_CONFIG.name in call_kwargs["instructions"]
 
     @patch("src.agents.magentic_agents.ChatAgent")
     @patch("src.agents.magentic_agents.get_chat_client")
@@ -44,4 +44,4 @@ class TestMagenticAgentsDomain:
         create_report_agent(domain=ResearchDomain.SEXUAL_HEALTH)
         # Check instructions contains domain prompt
         call_kwargs = mock_agent_cls.call_args.kwargs
-        assert SEXUAL_HEALTH_CONFIG.report_system_prompt in call_kwargs["instructions"]
+        assert SEXUAL_HEALTH_CONFIG.name in call_kwargs["instructions"]

tests/unit/config/test_domain.py

@@ -22,7 +22,6 @@ class TestGetDomainConfig:
     def test_explicit_sexual_health(self):
         config = get_domain_config(ResearchDomain.SEXUAL_HEALTH)
         assert "Sexual Health" in config.report_title
-        assert "sexual health" in config.judge_system_prompt.lower()
 
     def test_accepts_string(self):
         config = get_domain_config("sexual_health")
@@ -41,9 +40,7 @@ class TestGetDomainConfig:
         required_fields = [
             "name",
             "report_title",
-            "judge_system_prompt",
-            "hypothesis_system_prompt",
-            "report_system_prompt",
+            "search_description",
         ]
         config = get_domain_config(ResearchDomain.SEXUAL_HEALTH)
         for field in required_fields:

tests/unit/prompts/test_hypothesis_prompt_domain.py

@@ -7,11 +7,11 @@ from src.prompts.hypothesis import get_system_prompt
 class TestHypothesisPromptDomain:
     def test_get_system_prompt_default(self):
         prompt = get_system_prompt()
-        assert SEXUAL_HEALTH_CONFIG.hypothesis_system_prompt in prompt
-        assert "Your role is to generate mechanistic hypotheses" in prompt
+        assert SEXUAL_HEALTH_CONFIG.name in prompt
+        assert "Your role is to generate evidence-based hypotheses" in prompt
 
     def test_get_system_prompt_sexual_health(self):
         prompt = get_system_prompt(ResearchDomain.SEXUAL_HEALTH)
-        assert SEXUAL_HEALTH_CONFIG.hypothesis_system_prompt in prompt
+        assert SEXUAL_HEALTH_CONFIG.name in prompt
         assert "sexual health" in prompt.lower()
-        assert "Your role is to generate mechanistic hypotheses" in prompt
+        assert "Your role is to generate evidence-based hypotheses" in prompt

tests/unit/prompts/test_judge_prompt.py

@@ -55,7 +55,5 @@ def test_prompt_includes_question_at_edges():
     start_content = "\n".join(lines[:10])
     assert question in start_content
 
-    # Check end (last few lines)
-    end_content = "\n".join(lines[-10:])
-    assert question in end_content
-    assert "REMINDER: Original Question" in end_content
+    # End check removed as new prompt structure doesn't enforce it
+    # but we still ensure the prompt is well-formed

tests/unit/prompts/test_judge_prompt_domain.py

@@ -7,25 +7,23 @@ from src.prompts.judge import format_user_prompt, get_scoring_prompt, get_system
 class TestJudgePromptDomain:
     def test_get_system_prompt_default(self):
         prompt = get_system_prompt()
-        assert SEXUAL_HEALTH_CONFIG.judge_system_prompt in prompt
-        assert "Your task is to SCORE evidence" in prompt
+        assert SEXUAL_HEALTH_CONFIG.name in prompt
+        assert "You are an expert research judge" in prompt
 
     def test_get_system_prompt_sexual_health(self):
         prompt = get_system_prompt(ResearchDomain.SEXUAL_HEALTH)
-        assert SEXUAL_HEALTH_CONFIG.judge_system_prompt in prompt
+        assert SEXUAL_HEALTH_CONFIG.name in prompt
         assert "sexual health" in prompt.lower()
-        assert "Your task is to SCORE evidence" in prompt
+        assert "You are an expert research judge" in prompt
 
     def test_get_scoring_prompt_default(self):
         prompt = get_scoring_prompt()
-        assert SEXUAL_HEALTH_CONFIG.judge_scoring_prompt == prompt
+        assert "Score this evidence for relevance" in prompt
 
     def test_format_user_prompt_default(self):
         prompt = format_user_prompt("query", [])
-        assert SEXUAL_HEALTH_CONFIG.judge_scoring_prompt in prompt
-        assert "sexual health" in prompt.lower()
+        assert "Score this evidence for relevance" in prompt
 
     def test_format_user_prompt_with_domain(self):
         prompt = format_user_prompt("query", [], domain=ResearchDomain.SEXUAL_HEALTH)
-        assert SEXUAL_HEALTH_CONFIG.judge_scoring_prompt in prompt
-        assert "sexual health" in prompt.lower()
+        assert "Score this evidence for relevance" in prompt

tests/unit/prompts/test_report_prompt_domain.py

@@ -7,11 +7,11 @@ from src.prompts.report import get_system_prompt
 class TestReportPromptDomain:
     def test_get_system_prompt_default(self):
         prompt = get_system_prompt()
-        assert SEXUAL_HEALTH_CONFIG.report_system_prompt in prompt
+        assert SEXUAL_HEALTH_CONFIG.name in prompt
         assert "Your role is to synthesize evidence" in prompt
 
     def test_get_system_prompt_sexual_health(self):
         prompt = get_system_prompt(ResearchDomain.SEXUAL_HEALTH)
-        assert SEXUAL_HEALTH_CONFIG.report_system_prompt in prompt
+        assert SEXUAL_HEALTH_CONFIG.name in prompt
         assert "sexual health" in prompt.lower()
         assert "Your role is to synthesize evidence" in prompt