Make search limit configurable

docs/configuration/index.md

@@ -99,6 +99,7 @@ research:
   max_concurrency: 1
 
 search:
+  limit: 5                     # Default number of results to return
   vector_index_metric: cosine  # cosine, l2, or dot
   vector_refine_factor: 30
 

docs/configuration/storage.md

@@ -57,16 +57,18 @@ haiku.rag intelligently handles database creation based on operation type:
 
 This prevents the common mistake where a search query accidentally creates an empty database. To initialize your database, simply add your first document using `haiku-rag add` or `haiku-rag add-src`.
 
-## Vector Indexing
+## Search Settings
 
-Configure vector indexing behavior for efficient similarity search:
+Configure search behavior:
 
 ```yaml
 search:
+  limit: 5                     # Default number of results to return
   vector_index_metric: cosine  # cosine, l2, or dot
   vector_refine_factor: 30     # Re-ranking factor for accuracy
 ```
 
+- **limit**: Default number of search results to return when no limit is specified. Used by CLI, MCP server, and API. Default: 5
 - **vector_index_metric**: Distance metric for vector similarity:
   - `cosine`: Cosine similarity (default, best for most embeddings)
   - `l2`: Euclidean distance

haiku_rag_slim/haiku/rag/app.py

@@ -265,7 +265,9 @@ class HaikuRAGApp:
                     f"[yellow]Document with id {doc_id} not found.[/yellow]"
                 )
 
-    async def search(self, query: str, limit: int = 5, filter: str | None = None):
+    async def search(
+        self, query: str, limit: int | None = None, filter: str | None = None
+    ):
         async with HaikuRAG(db_path=self.db_path, config=self.config) as self.client:
             results = await self.client.search(query, limit=limit, filter=filter)
             if not results:

haiku_rag_slim/haiku/rag/cli.py

@@ -239,11 +239,11 @@ def search(
     query: str = typer.Argument(
         help="The search query to use",
     ),
-    limit: int = typer.Option(
-        5,
+    limit: int | None = typer.Option(
+        None,
         "--limit",
         "-l",
-        help="Maximum number of results to return",
+        help="Maximum number of results to return (default: config search.default_limit)",
     ),
     filter: str | None = typer.Option(
         None,

haiku_rag_slim/haiku/rag/client.py

@@ -834,7 +834,7 @@ class HaikuRAG:
     async def search(
         self,
         query: str,
-        limit: int = 5,
+        limit: int | None = None,
         search_type: str = "hybrid",
         filter: str | None = None,
     ) -> list[SearchResult]:
@@ -842,13 +842,16 @@ class HaikuRAG:
 
         Args:
             query: The search query string.
-            limit: Maximum number of results to return.
+            limit: Maximum number of results to return. Defaults to config.search.default_limit.
             search_type: Type of search - "vector", "fts", or "hybrid" (default).
             filter: Optional SQL WHERE clause to filter documents before searching chunks.
 
         Returns:
             List of SearchResult objects ordered by relevance.
         """
+        if limit is None:
+            limit = self._config.search.limit
+
         reranker = get_reranker(config=self._config)
 
         if reranker is None:

haiku_rag_slim/haiku/rag/config/models.py

@@ -123,6 +123,7 @@ class ProcessingConfig(BaseModel):
 
 
 class SearchConfig(BaseModel):
+    limit: int = 5
     vector_index_metric: Literal["cosine", "l2", "dot"] = "cosine"
     vector_refine_factor: int = 30
 

haiku_rag_slim/haiku/rag/graph/common/nodes.py

@@ -101,7 +101,7 @@ def create_plan_node[AgentDepsT: GraphAgentDeps](
 
             @plan_agent.tool
             async def gather_context(
-                ctx2: RunContext[AgentDepsT], query: str, limit: int = 6
+                ctx2: RunContext[AgentDepsT], query: str, limit: int | None = None
             ) -> str:
                 results = await ctx2.deps.client.search(query, limit=limit)
                 results = await ctx2.deps.client.expand_context(results)
@@ -227,7 +227,7 @@ async def _do_search[AgentDepsT: GraphAgentDeps](
 
     @agent.tool
     async def search_and_answer(
-        ctx2: RunContext[AgentDepsT], query: str, limit: int = 5
+        ctx2: RunContext[AgentDepsT], query: str, limit: int | None = None
     ) -> str:
         """Search the knowledge base for relevant documents.
 

haiku_rag_slim/haiku/rag/mcp.py

@@ -79,11 +79,13 @@ def create_mcp_server(db_path: Path, config: AppConfig = Config) -> FastMCP:
             return None
 
     @mcp.tool()
-    async def search_documents(query: str, limit: int = 5) -> list[SearchResult]:
+    async def search_documents(
+        query: str, limit: int | None = None
+    ) -> list[SearchResult]:
         """Search the RAG system for documents using hybrid search (vector similarity + full-text search)."""
         try:
             async with HaikuRAG(db_path, config=config) as rag:
-                return await rag.search(query, limit)
+                return await rag.search(query, limit=limit)
         except Exception:
             return []
 

haiku_rag_slim/haiku/rag/qa/agent.py

@@ -39,7 +39,7 @@ class QuestionAnswerAgent:
         async def search_documents(
             ctx: RunContext[Dependencies],
             query: str,
-            limit: int = 5,
+            limit: int | None = None,
         ) -> str:
             """Search the knowledge base for relevant documents.
 

tests/test_app.py

@@ -159,7 +159,7 @@ async def test_search(app: HaikuRAGApp, monkeypatch):
     with patch("haiku.rag.app.HaikuRAG", return_value=mock_client):
         await app.search("query")
 
-    mock_client.search.assert_called_once_with("query", limit=5, filter=None)
+    mock_client.search.assert_called_once_with("query", limit=None, filter=None)
     assert mock_rich_print_search.call_count == len(mock_results)
 
 
@@ -176,7 +176,7 @@ async def test_search_no_results(app: HaikuRAGApp, monkeypatch):
     with patch("haiku.rag.app.HaikuRAG", return_value=mock_client):
         await app.search("query")
 
-    mock_client.search.assert_called_once_with("query", limit=5, filter=None)
+    mock_client.search.assert_called_once_with("query", limit=None, filter=None)
     mock_print.assert_called_once_with("[yellow]No results found.[/yellow]")
 
 

tests/test_cli.py

@@ -171,7 +171,7 @@ def test_search():
 
         assert result.exit_code == 0
         mock_app_instance.search.assert_called_once_with(
-            query="query", limit=5, filter=None
+            query="query", limit=None, filter=None
         )
 
 

tests/test_mcp.py

@@ -148,7 +148,7 @@ async def test_mcp_search_documents():
             assert result[0].content == "Result 1"
             assert result[0].score == 0.9
             assert result[1].document_id == "doc2"
-            mock_rag.search.assert_called_once_with("test query", 5)
+            mock_rag.search.assert_called_once_with("test query", limit=5)
 
 
 @pytest.mark.asyncio

tests/test_search.py

@@ -225,8 +225,9 @@ async def test_search_graceful_degradation(temp_db_path):
     custom_chunks = [
         Chunk(content="Custom chunk without docling metadata", metadata={}),
     ]
+    docling_doc = await client.convert("Document with custom chunks")
     await client.import_document(
-        content="Document with custom chunks",
+        docling_document=docling_doc,
         chunks=custom_chunks,
         uri="https://example.com/custom.html",
     )