updates

Author: willemcdejongh

agent-os/customize/os/filter_knowledge.mdx

@@ -1,5 +1,5 @@
 ---
-title: Knowledge Filtering
+title: Filter Knowledge
 description: Learn how to use advanced filter expressions through the Agno API for precise knowledge base filtering.
 keywords: [api, filtering, filterexpr, knowledge, search, rest api]
 ---
@@ -272,93 +272,6 @@ def validate_and_send_filter(filter_expr, message):
         return None
 ```
 
-## Complete Examples
-
-### Example 1: Simple Dict Filter
-
-For straightforward filtering with AND logic:
-
-```python
-import requests
-import json
-
-# Define your agent's API endpoint
-AGENT_URL = "http://localhost:7777/agents/agno-knowledge-agent/runs"
-
-# Create simple dict filter
-filter_dict = {
-    "docs": "agno",
-    "category": "features",
-    "status": "published"
-}
-
-# Serialize filter
-filter_json = json.dumps(filter_dict)
-
-# Send request
-response = requests.post(
-    AGENT_URL,
-    data={
-        "message": "What are agno's key features?",
-        "stream": "false",
-        "knowledge_filters": filter_json,
-        "session_id": "user-123",
-        "user_id": "demo-user",
-    }
-)
-
-# Process response
-if response.ok:
-    result = response.json()
-    print(f"Agent response: {result.get('content')}")
-    print(f"Documents used: {len(result.get('documents', []))}")
-else:
-    print(f"Error: {response.status_code} - {response.text}")
-```
-
-### Example 2: Advanced Filter Expression
-
-For complex filtering with multiple logical operators:
-
-```python
-import requests
-import json
-from agno.filters import EQ, GT, AND, IN, NOT
-
-# Define your agent's API endpoint
-AGENT_URL = "http://localhost:7777/agents/tech-docs-agent/runs"
-
-# Build a complex filter
-filter_expr = AND(
-    EQ("status", "published"),
-    IN("category", ["tutorial", "guide"]),
-    GT("word_count", 500),
-    NOT(EQ("difficulty", "beginner"))
-)
-
-# Serialize filter
-filter_json = json.dumps(filter_expr.to_dict())
-
-# Send request
-response = requests.post(
-    AGENT_URL,
-    data={
-        "message": "Show me advanced tutorials on deployment",
-        "stream": "false",
-        "knowledge_filters": filter_json,
-        "session_id": "user-123-session",
-    }
-)
-
-# Process response
-if response.ok:
-    result = response.json()
-    print(f"Agent response: {result.get('content')}")
-    print(f"Documents used: {len(result.get('documents', []))}")
-else:
-    print(f"Error: {response.status_code} - {response.text}")
-```
-
 ## Next Steps
 
 <CardGroup cols={2}>
@@ -390,5 +303,12 @@ else:
   >
     Optimize your search strategies
   </Card>
+  <Card
+    title="Filter Expressions Example"
+    icon="code"
+    href="/examples/concepts/knowledge/filters/filter-expressions"
+  >
+    See working examples of filter expressions via API
+  </Card>
 </CardGroup>
 

concepts/knowledge/core-concepts/advanced-filtering.mdx

@@ -419,15 +419,7 @@ def safe_filter_search(agent, query, filters):
 
 ### Vector Database Support
 
-Advanced filter expressions (using `FilterExpr` like `EQ()`, `AND()`, etc.) have varying support across vector databases:
-
-| Vector DB | Dictionary Filters `{"key": "value"}` | FilterExpr `EQ()`, `AND()`, etc. |
-|-----------|--------------------------------------|----------------------------------|
-| **PgVector** | ✅ Full support | ✅ Full support (native) |
-| **LanceDB** | ✅ Post-search filtering | ⚠️ Not yet supported* |
-| **ChromaDB** | ✅ Full support | ⚠️ Not yet supported |
-| **Pinecone** | ✅ Full support | ⚠️ Not yet supported |
-| **Qdrant** | ✅ Full support | ⚠️ Not yet supported |
+Advanced filter expressions (using `FilterExpr` like `EQ()`, `AND()`, etc.) is currently only supported in PGVector.
 
 <Note>
 **What happens with unsupported FilterExpr:**
@@ -447,23 +439,17 @@ knowledge_filters=[AND(EQ("department", "hr"), EQ("year", 2024))]
 ```
 </Note>
 
-**Coming Soon:**
-- 🔄 LanceDB FilterExpr support (in development)
-- 🔄 ChromaDB FilterExpr support (planned)  
-- 🔄 Pinecone FilterExpr support (planned)
-- 🔄 Universal FilterExpr support across all vector databases
-
 ### Agentic Filtering Compatibility
 
 Advanced filter expressions (`FilterExpr`) are **not compatible with agentic filtering**, where agents dynamically construct filters based on conversation context.
 
 **For agentic filtering, use dictionary format:**
 
 ```python
-# ✅ Works with agentic filtering (agent decides filters dynamically)
+# Works with agentic filtering (agent decides filters dynamically)
 knowledge_filters = [{"department": "hr", "document_type": "policy"}]
 
-# ❌ Does not work with agentic filtering (static, predefined logic)
+# Does not work with agentic filtering (static, predefined logic)
 knowledge_filters = [AND(EQ("department", "hr"), EQ("document_type", "policy"))]
 ```
 
@@ -474,28 +460,6 @@ knowledge_filters = [AND(EQ("department", "hr"), EQ("document_type", "policy"))]
 | **Dictionary format** | Agent dynamically chooses filters based on conversation | User mentions "HR policies" → agent adds `{"department": "hr"}` |
 | **Filter expressions** | You need complex, predetermined logic with full control | Always exclude drafts AND filter by multiple regions with OR logic |
 
-**Hybrid Approach:**
-
-If you need both agentic decision-making and complex filtering:
-
-```python
-from agno.filters import AND, EQ, GT, NOT
-
-# Define filter presets
-FILTER_PRESETS = {
-    "recent_hr": [AND(EQ("department", "hr"), GT("year", 2023))],
-    "active_engineering": [AND(EQ("department", "engineering"), NOT(EQ("status", "archived")))],
-    "all_public": [EQ("visibility", "public")]
-}
-
-# Agent chooses which preset to use based on context
-def get_filter_for_context(user_query: str):
-    if "recent" in user_query.lower() and "hr" in user_query.lower():
-        return FILTER_PRESETS["recent_hr"]
-    # ... more logic
-    return None
-```
-
 
 ## Using Filters Through the API
 

docs.json

@@ -1703,6 +1703,7 @@
                     "pages": [
                       "examples/concepts/knowledge/filters/agentic-filtering",
                       "examples/concepts/knowledge/filters/async-filtering",
+                      "examples/concepts/knowledge/filters/filter-expressions",
                       "examples/concepts/knowledge/filters/filtering",
                       "examples/concepts/knowledge/filters/filtering_on_load",
                       "examples/concepts/knowledge/filters/filtering_with_invalid_keys",