👌 IMPROVE: Primitives

Author: arre-ankit

.pre-commit-config.yaml

@@ -24,17 +24,3 @@ repos:
     hooks:
       - id: isort
         args: [--profile=black, --line-length=88]
-
-  - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.1.9
-    hooks:
-      - id: ruff
-        args: [--fix]
-
-  - repo: https://github.com/pre-commit/mirrors-mypy
-    rev: v1.8.0
-    hooks:
-      - id: mypy
-        args: [--strict, --ignore-missing-imports]
-        additional_dependencies: [types-requests>=2.28.0]
-        exclude: ^(tests/|examples/)

README.md

@@ -98,16 +98,16 @@ response = lb.pipes.run(
 runner = get_typed_runner(response)
 
 # Register event handlers
-runner.on(StreamEventType.CONNECT, lambda e: 
+runner.on(StreamEventType.CONNECT, lambda e:
     print(f"✓ Connected to thread: {e['threadId']}"))
 
-runner.on(StreamEventType.CONTENT, lambda e: 
+runner.on(StreamEventType.CONTENT, lambda e:
     print(e["content"], end="", flush=True))
 
-runner.on(StreamEventType.TOOL_CALL, lambda e: 
+runner.on(StreamEventType.TOOL_CALL, lambda e:
     print(f"\n🔧 Tool: {e['toolCall']['function']['name']}"))
 
-runner.on(StreamEventType.END, lambda e: 
+runner.on(StreamEventType.END, lambda e:
     print(f"\n⏱️  Duration: {e['duration']:.2f}s"))
 
 # Process the stream
@@ -160,7 +160,7 @@ results = lb.memories.retrieve(
 
 ```python
 # Run an agent with tools
-response = lb.agent_run(
+response = lb.agent.run(
     model="openai:gpt-4",
     messages=[{"role": "user", "content": "Search for AI news"}],
     tools=[{"type": "function", "function": {...}}],

examples/agent/README.md

@@ -140,7 +140,7 @@ langbase = Langbase(api_key=os.environ.get("LANGBASE_API_KEY"))
 Basic agent run pattern:
 
 ```python
-response = langbase.agent_run(
+response = langbase.agent.run(
     model="openai:gpt-4.1-mini",
     api_key=os.environ.get("LLM_API_KEY"),
     instructions="Your instructions here",
@@ -184,4 +184,4 @@ logging.basicConfig(level=logging.DEBUG)
 - Explore the [Langbase Documentation](https://docs.langbase.com)
 - Try creating your own custom tools
 - Experiment with different models and parameters
-- Build multi-agent workflows 
+- Build multi-agent workflows

examples/agent/agent.run.mcp.py

@@ -30,7 +30,7 @@ def main():
     langbase = Langbase(api_key=langbase_api_key)
 
     # Run the agent with MCP server
-    response = langbase.agent_run(
+    response = langbase.agent.run(
         stream=False,
         model="openai:gpt-4.1-mini",
         api_key=llm_api_key,

examples/agent/agent.run.memory.py

@@ -39,7 +39,7 @@ def main():
     )
 
     # Step 2: Run the agent with the retrieved memory
-    response = langbase.agent_run(
+    response = langbase.agent.run(
         model="openai:gpt-4.1",
         api_key=llm_api_key,
         instructions="You are a career advisor who helps users understand AI job roles.",

examples/agent/agent.run.py

@@ -32,7 +32,7 @@ def main():
     langbase = Langbase(api_key=langbase_api_key)
 
     # Run the agent
-    response = langbase.agent_run(
+    response = langbase.agent.run(
         stream=False,
         model="openai:gpt-4.1-mini",
         api_key=llm_api_key,

examples/agent/agent.run.stream.py

@@ -28,7 +28,7 @@ def main():
 
     try:
         # Get readable stream - equivalent to const {stream} = await langbase.agent.run(...)
-        response = langbase.agent_run(
+        response = langbase.agent.run(
             stream=True,
             model="openai:gpt-4.1-mini",
             instructions="You are a helpful assistant that help users summarize text.",

examples/agent/agent.run.structured.py

@@ -51,7 +51,7 @@ def main():
     }
 
     # Run the agent with structured output
-    response = langbase.agent_run(
+    response = langbase.agent.run(
         model="openai:gpt-4.1",
         api_key=llm_api_key,
         instructions="You are a helpful math tutor. Guide the user through the solution step by step.",

examples/agent/agent.run.tool.py

@@ -49,7 +49,7 @@ def send_email(args):
     response = requests.post(
         "https://api.resend.com/emails",
         headers={
-            "Authorization": f'Bearer {os.environ.get("RESEND_API_KEY")}',
+            "Authorization": f"Bearer {os.environ.get('RESEND_API_KEY')}",
             "Content-Type": "application/json",
         },
         json={
@@ -99,7 +99,7 @@ def main():
     input_messages = [{"role": "user", "content": "Send a welcome email to Sam."}]
 
     # Initial run with tool
-    response = langbase.agent_run(
+    response = langbase.agent.run(
         model="openai:gpt-4.1-mini",
         api_key=llm_api_key,
         instructions="You are an email sending assistant.",
@@ -159,7 +159,7 @@ def main():
                 continue
 
     # Final agent response with tool result
-    final_response = langbase.agent_run(
+    final_response = langbase.agent.run(
         model="openai:gpt-4.1-mini",
         api_key=os.environ.get("OPENAI_API_KEY"),
         instructions="You are an email sending assistant. Confirm the email has been sent successfully.",

examples/agent/agent.run.workflow.py

@@ -96,8 +96,7 @@ async def flaky_operation():
                 ],
             )
             return response["completion"]
-        else:
-            raise Exception("Temporary service unavailable")
+        raise Exception("Temporary service unavailable")
 
     try:
         analysis = await workflow.step(