diff --git a/CHANGELOG.md b/CHANGELOG.md
index 0e7445ac..d282aec5 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,6 +1,10 @@
 # Changelog
 ## [Unreleased]
 
+### Fixed
+
+- **Search tool regression**: Removed LLM-facing `filter` parameter from search and list_documents tools. The SQL WHERE clause description confused LLMs, degrading QA accuracy. Document filtering is now handled programmatically via `base_filter` and `state.document_filter`
+
 ## [0.32.2] - 2026-02-28
 
 ### Fixed
diff --git a/haiku_rag_slim/haiku/rag/skills/rag.py b/haiku_rag_slim/haiku/rag/skills/rag.py
index 776db28a..856b6b30 100644
--- a/haiku_rag_slim/haiku/rag/skills/rag.py
+++ b/haiku_rag_slim/haiku/rag/skills/rag.py
@@ -140,19 +140,17 @@ def create_skill(
         ctx: RunContext[SkillRunDeps],
         limit: int | None = None,
         offset: int | None = None,
-        filter: str | None = None,
     ) -> list[dict[str, Any]]:
-        """List documents in the knowledge base with optional pagination and filtering.
+        """List documents in the knowledge base with optional pagination.
 
         Args:
             limit: Maximum number of documents to return.
             offset: Number of documents to skip.
-            filter: Optional SQL WHERE clause to filter documents.
         """
         from haiku.rag.client import HaikuRAG
 
         async with HaikuRAG(db_path, config=config, read_only=True) as rag:
-            documents = await rag.list_documents(limit, offset, filter)
+            documents = await rag.list_documents(limit, offset)
             result = [
                 {
                     "id": doc.id,
diff --git a/tests/tools/test_search.py b/tests/tools/test_search.py
index 69ee447f..33adc1d6 100644
--- a/tests/tools/test_search.py
+++ b/tests/tools/test_search.py
@@ -67,19 +67,6 @@ class TestSearchToolExecution:
 
             assert result == "No results found."
 
-    @pytest.mark.asyncio
-    async def test_search_with_filter(self, search_client, search_config):
-        """Search tool respects filter parameter."""
-        accumulated: list[SearchResult] = []
-        toolset = create_search_toolset(search_config, on_results=accumulated.extend)
-
-        search_tool = toolset.tools["search"]
-        ctx = make_ctx(search_client)
-        await search_tool.function(ctx, "programming", filter="title LIKE '%Python%'")
-
-        for r in accumulated:
-            assert "JavaScript" not in (r.document_title or "")
-
     @pytest.mark.asyncio
     async def test_search_with_base_filter(self, search_client, search_config):
         """Search toolset respects base_filter parameter."""