letta-nightly 0.6.37.dev20250310103931__tar.gz → 0.6.38.dev20250312104155__tar.gz

{letta_nightly-0.6.37.dev20250310103931 → letta_nightly-0.6.38.dev20250312104155}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: letta-nightly
-Version: 0.6.37.dev20250310103931
+Version: 0.6.38.dev20250312104155
 Summary: Create LLM agents with long-term memory and custom tools
 License: Apache License
 Author: Letta Team
@@ -14,6 +14,7 @@ Classifier: Programming Language :: Python :: 3.12
 Provides-Extra: all
 Provides-Extra: bedrock
 Provides-Extra: cloud-tool-sandbox
+Provides-Extra: desktop
 Provides-Extra: dev
 Provides-Extra: external-tools
 Provides-Extra: google
@@ -28,17 +29,17 @@ Requires-Dist: black[jupyter] (>=24.2.0,<25.0.0) ; extra == "dev" or extra == "a
 Requires-Dist: boto3 (>=1.36.24,<2.0.0) ; extra == "bedrock"
 Requires-Dist: brotli (>=1.1.0,<2.0.0)
 Requires-Dist: colorama (>=0.4.6,<0.5.0)
-Requires-Dist: composio-core (>=0.7.2,<0.8.0)
-Requires-Dist: composio-langchain (>=0.7.2,<0.8.0)
-Requires-Dist: datamodel-code-generator[http] (>=0.25.0,<0.26.0)
-Requires-Dist: datasets (>=2.14.6,<3.0.0) ; extra == "dev" or extra == "all"
+Requires-Dist: composio-core (>=0.7.7,<0.8.0)
+Requires-Dist: composio-langchain (>=0.7.7,<0.8.0)
+Requires-Dist: datamodel-code-generator[http] (>=0.25.0,<0.26.0) ; extra == "desktop" or extra == "all"
+Requires-Dist: datasets (>=2.14.6,<3.0.0) ; extra == "dev" or extra == "desktop" or extra == "all"
 Requires-Dist: demjson3 (>=3.0.6,<4.0.0)
-Requires-Dist: docker (>=7.1.0,<8.0.0) ; extra == "external-tools" or extra == "all"
+Requires-Dist: docker (>=7.1.0,<8.0.0) ; extra == "external-tools" or extra == "desktop" or extra == "all"
 Requires-Dist: docstring-parser (>=0.16,<0.17)
 Requires-Dist: docx2txt (>=0.8,<0.9)
 Requires-Dist: e2b-code-interpreter (>=1.0.3,<2.0.0) ; extra == "cloud-tool-sandbox"
 Requires-Dist: faker (>=36.1.0,<37.0.0)
-Requires-Dist: fastapi (>=0.115.6,<0.116.0) ; extra == "server" or extra == "all"
+Requires-Dist: fastapi (>=0.115.6,<0.116.0) ; extra == "server" or extra == "desktop" or extra == "all"
 Requires-Dist: google-genai (>=1.1.0,<2.0.0) ; extra == "google"
 Requires-Dist: grpcio (>=1.68.1,<2.0.0)
 Requires-Dist: grpcio-tools (>=1.68.1,<2.0.0)
@@ -47,12 +48,12 @@ Requires-Dist: httpx (>=0.28.0,<0.29.0)
 Requires-Dist: httpx-sse (>=0.4.0,<0.5.0)
 Requires-Dist: isort (>=5.13.2,<6.0.0) ; extra == "dev" or extra == "all"
 Requires-Dist: jinja2 (>=3.1.5,<4.0.0)
-Requires-Dist: langchain (>=0.3.7,<0.4.0) ; extra == "external-tools" or extra == "all"
-Requires-Dist: langchain-community (>=0.3.7,<0.4.0) ; extra == "external-tools" or extra == "all"
+Requires-Dist: langchain (>=0.3.7,<0.4.0) ; extra == "external-tools" or extra == "desktop" or extra == "all"
+Requires-Dist: langchain-community (>=0.3.7,<0.4.0) ; extra == "external-tools" or extra == "desktop" or extra == "all"
 Requires-Dist: letta_client (>=0.1.54,<0.2.0)
 Requires-Dist: llama-index (>=0.12.2,<0.13.0)
 Requires-Dist: llama-index-embeddings-openai (>=0.3.1,<0.4.0)
-Requires-Dist: locust (>=2.31.5,<3.0.0) ; extra == "dev" or extra == "all"
+Requires-Dist: locust (>=2.31.5,<3.0.0) ; extra == "dev" or extra == "desktop" or extra == "all"
 Requires-Dist: marshmallow-sqlalchemy (>=1.4.1,<2.0.0)
 Requires-Dist: nltk (>=3.8.1,<4.0.0)
 Requires-Dist: numpy (>=1.26.2,<2.0.0)
@@ -63,16 +64,16 @@ Requires-Dist: opentelemetry-instrumentation-requests (==0.51b0)
 Requires-Dist: opentelemetry-sdk (==1.30.0)
 Requires-Dist: pathvalidate (>=3.2.1,<4.0.0)
 Requires-Dist: pexpect (>=4.9.0,<5.0.0) ; extra == "dev" or extra == "all"
-Requires-Dist: pg8000 (>=1.30.3,<2.0.0) ; extra == "postgres" or extra == "all"
-Requires-Dist: pgvector (>=0.2.3,<0.3.0) ; extra == "postgres" or extra == "all"
+Requires-Dist: pg8000 (>=1.30.3,<2.0.0) ; extra == "postgres" or extra == "desktop" or extra == "all"
+Requires-Dist: pgvector (>=0.2.3,<0.3.0) ; extra == "postgres" or extra == "desktop" or extra == "all"
 Requires-Dist: pre-commit (>=3.5.0,<4.0.0) ; extra == "dev" or extra == "all"
 Requires-Dist: prettytable (>=3.9.0,<4.0.0)
-Requires-Dist: psycopg2 (>=2.9.10,<3.0.0) ; extra == "postgres" or extra == "all"
-Requires-Dist: psycopg2-binary (>=2.9.10,<3.0.0) ; extra == "postgres" or extra == "all"
+Requires-Dist: psycopg2 (>=2.9.10,<3.0.0) ; extra == "postgres" or extra == "desktop" or extra == "all"
+Requires-Dist: psycopg2-binary (>=2.9.10,<3.0.0) ; extra == "postgres" or extra == "desktop" or extra == "all"
 Requires-Dist: pydantic (>=2.7.4,<2.10.0)
 Requires-Dist: pydantic-settings (>=2.2.1,<3.0.0)
 Requires-Dist: pyhumps (>=3.8.0,<4.0.0)
-Requires-Dist: pyright (>=1.1.347,<2.0.0) ; extra == "dev" or extra == "all"
+Requires-Dist: pyright (>=1.1.347,<2.0.0) ; extra == "dev" or extra == "desktop" or extra == "all"
 Requires-Dist: pytest-asyncio (>=0.23.2,<0.24.0) ; extra == "dev" or extra == "all"
 Requires-Dist: pytest-order (>=1.2.0,<2.0.0) ; extra == "dev" or extra == "all"
 Requires-Dist: python-box (>=7.1.1,<8.0.0)
@@ -90,8 +91,8 @@ Requires-Dist: sqlalchemy-utils (>=0.41.2,<0.42.0)
 Requires-Dist: sqlmodel (>=0.0.16,<0.0.17)
 Requires-Dist: tqdm (>=4.66.1,<5.0.0)
 Requires-Dist: typer[all] (>=0.9.0,<0.10.0)
-Requires-Dist: uvicorn (>=0.24.0.post1,<0.25.0) ; extra == "server" or extra == "all"
-Requires-Dist: wikipedia (>=1.4.0,<2.0.0) ; extra == "external-tools" or extra == "tests" or extra == "all"
+Requires-Dist: uvicorn (>=0.24.0.post1,<0.25.0) ; extra == "server" or extra == "desktop" or extra == "all"
+Requires-Dist: wikipedia (>=1.4.0,<2.0.0) ; extra == "external-tools" or extra == "tests" or extra == "desktop" or extra == "all"
 Description-Content-Type: text/markdown
 
 <p align="center">

{letta_nightly-0.6.37.dev20250310103931 → letta_nightly-0.6.38.dev20250312104155}/letta/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "0.6.37"
+__version__ = "0.6.38"
 
 # import clients
 from letta.client.client import LocalClient, RESTClient, create_client

{letta_nightly-0.6.37.dev20250310103931 → letta_nightly-0.6.38.dev20250312104155}/letta/agent.py

@@ -29,6 +29,7 @@ from letta.helpers.json_helpers import json_dumps, json_loads
 from letta.interface import AgentInterface
 from letta.llm_api.helpers import calculate_summarizer_cutoff, get_token_counts_for_messages, is_context_overflow_error
 from letta.llm_api.llm_api_tools import create
+from letta.llm_api.llm_client import LLMClient
 from letta.local_llm.utils import num_tokens_from_functions, num_tokens_from_messages
 from letta.log import get_logger
 from letta.memory import summarize_messages
@@ -356,19 +357,38 @@ class Agent(BaseAgent):
         for attempt in range(1, empty_response_retry_limit + 1):
             try:
                 log_telemetry(self.logger, "_get_ai_reply create start")
-                response = create(
+                # New LLM client flow
+                llm_client = LLMClient.create(
+                    agent_id=self.agent_state.id,
                     llm_config=self.agent_state.llm_config,
-                    messages=message_sequence,
-                    user_id=self.agent_state.created_by_id,
-                    functions=allowed_functions,
-                    # functions_python=self.functions_python, do we need this?
-                    function_call=function_call,
-                    first_message=first_message,
-                    force_tool_call=force_tool_call,
-                    stream=stream,
-                    stream_interface=self.interface,
                     put_inner_thoughts_first=put_inner_thoughts_first,
+                    actor_id=self.agent_state.created_by_id,
                 )
+
+                if llm_client and not stream:
+                    response = llm_client.send_llm_request(
+                        messages=message_sequence,
+                        tools=allowed_functions,
+                        tool_call=function_call,
+                        stream=stream,
+                        first_message=first_message,
+                        force_tool_call=force_tool_call,
+                    )
+                else:
+                    # Fallback to existing flow
+                    response = create(
+                        llm_config=self.agent_state.llm_config,
+                        messages=message_sequence,
+                        user_id=self.agent_state.created_by_id,
+                        functions=allowed_functions,
+                        # functions_python=self.functions_python, do we need this?
+                        function_call=function_call,
+                        first_message=first_message,
+                        force_tool_call=force_tool_call,
+                        stream=stream,
+                        stream_interface=self.interface,
+                        put_inner_thoughts_first=put_inner_thoughts_first,
+                    )
                 log_telemetry(self.logger, "_get_ai_reply create finish")
 
                 # These bottom two are retryable
@@ -632,7 +652,7 @@ class Agent(BaseAgent):
                     function_args,
                     function_response,
                     messages,
-                    [tool_return] if tool_return else None,
+                    [tool_return],
                     include_function_failed_message=True,
                 )
                 return messages, False, True  # force a heartbeat to allow agent to handle error
@@ -659,7 +679,7 @@ class Agent(BaseAgent):
                         "content": function_response,
                         "tool_call_id": tool_call_id,
                     },
-                    tool_returns=[tool_return] if tool_return else None,
+                    tool_returns=[tool_return] if sandbox_run_result else None,
                 )
             )  # extend conversation with function response
             self.interface.function_message(f"Ran {function_name}({function_args})", msg_obj=messages[-1])
@@ -909,6 +929,7 @@ class Agent(BaseAgent):
             # Log step - this must happen before messages are persisted
             step = self.step_manager.log_step(
                 actor=self.user,
+                agent_id=self.agent_state.id,
                 provider_name=self.agent_state.llm_config.model_endpoint_type,
                 model=self.agent_state.llm_config.model,
                 model_endpoint=self.agent_state.llm_config.model_endpoint,
@@ -1174,6 +1195,7 @@ class Agent(BaseAgent):
             memory_edit_timestamp=get_utc_time(),
             previous_message_count=self.message_manager.size(actor=self.user, agent_id=self.agent_state.id),
             archival_memory_size=self.agent_manager.passage_size(actor=self.user, agent_id=self.agent_state.id),
+            recent_passages=self.agent_manager.list_passages(actor=self.user, agent_id=self.agent_state.id, ascending=False, limit=10),
         )
         num_tokens_external_memory_summary = count_tokens(external_memory_summary)
 

{letta_nightly-0.6.37.dev20250310103931 → letta_nightly-0.6.38.dev20250312104155}/letta/client/client.py

@@ -4,7 +4,6 @@ import time
 from typing import Callable, Dict, Generator, List, Optional, Union
 
 import requests
-from openai.types.chat.chat_completion_message_tool_call import ChatCompletionMessageToolCall as OpenAIToolCall
 
 import letta.utils
 from letta.constants import ADMIN_PREFIX, BASE_MEMORY_TOOLS, BASE_TOOLS, DEFAULT_HUMAN, DEFAULT_PERSONA, FUNCTION_RETURN_CHAR_LIMIT
@@ -29,7 +28,7 @@ from letta.schemas.letta_request import LettaRequest, LettaStreamingRequest
 from letta.schemas.letta_response import LettaResponse, LettaStreamingResponse
 from letta.schemas.llm_config import LLMConfig
 from letta.schemas.memory import ArchivalMemorySummary, ChatMemory, CreateArchivalMemory, Memory, RecallMemorySummary
-from letta.schemas.message import Message, MessageCreate, MessageUpdate
+from letta.schemas.message import Message, MessageCreate
 from letta.schemas.openai.chat_completion_response import UsageStatistics
 from letta.schemas.organization import Organization
 from letta.schemas.passage import Passage
@@ -640,30 +639,6 @@ class RESTClient(AbstractClient):
         # refresh and return agent
         return self.get_agent(agent_state.id)
 
-    def update_message(
-        self,
-        agent_id: str,
-        message_id: str,
-        role: Optional[MessageRole] = None,
-        text: Optional[str] = None,
-        name: Optional[str] = None,
-        tool_calls: Optional[List[OpenAIToolCall]] = None,
-        tool_call_id: Optional[str] = None,
-    ) -> Message:
-        request = MessageUpdate(
-            role=role,
-            content=text,
-            name=name,
-            tool_calls=tool_calls,
-            tool_call_id=tool_call_id,
-        )
-        response = requests.patch(
-            f"{self.base_url}/{self.api_prefix}/agents/{agent_id}/messages/{message_id}", json=request.model_dump(), headers=self.headers
-        )
-        if response.status_code != 200:
-            raise ValueError(f"Failed to update message: {response.text}")
-        return Message(**response.json())
-
     def update_agent(
         self,
         agent_id: str,
@@ -2436,30 +2411,6 @@ class LocalClient(AbstractClient):
         # TODO: get full agent state
         return self.server.agent_manager.get_agent_by_id(agent_state.id, actor=self.user)
 
-    def update_message(
-        self,
-        agent_id: str,
-        message_id: str,
-        role: Optional[MessageRole] = None,
-        text: Optional[str] = None,
-        name: Optional[str] = None,
-        tool_calls: Optional[List[OpenAIToolCall]] = None,
-        tool_call_id: Optional[str] = None,
-    ) -> Message:
-        message = self.server.update_agent_message(
-            agent_id=agent_id,
-            message_id=message_id,
-            request=MessageUpdate(
-                role=role,
-                content=text,
-                name=name,
-                tool_calls=tool_calls,
-                tool_call_id=tool_call_id,
-            ),
-            actor=self.user,
-        )
-        return message
-
     def update_agent(
         self,
         agent_id: str,

{letta_nightly-0.6.37.dev20250310103931 → letta_nightly-0.6.38.dev20250312104155}/letta/constants.py

@@ -50,7 +50,7 @@ BASE_TOOLS = ["send_message", "conversation_search", "archival_memory_insert", "
 # Base memory tools CAN be edited, and are added by default by the server
 BASE_MEMORY_TOOLS = ["core_memory_append", "core_memory_replace"]
 # Multi agent tools
-MULTI_AGENT_TOOLS = ["send_message_to_agent_and_wait_for_reply", "send_message_to_agents_matching_all_tags", "send_message_to_agent_async"]
+MULTI_AGENT_TOOLS = ["send_message_to_agent_and_wait_for_reply", "send_message_to_agents_matching_tags", "send_message_to_agent_async"]
 # Set of all built-in Letta tools
 LETTA_TOOL_SET = set(BASE_TOOLS + BASE_MEMORY_TOOLS + MULTI_AGENT_TOOLS)
 

{letta_nightly-0.6.37.dev20250310103931 → letta_nightly-0.6.38.dev20250312104155}/letta/functions/function_sets/multi_agent.py

@@ -2,7 +2,7 @@ import asyncio
 from typing import TYPE_CHECKING, List
 
 from letta.functions.helpers import (
-    _send_message_to_agents_matching_all_tags_async,
+    _send_message_to_agents_matching_tags_async,
     execute_send_message_to_agent,
     fire_and_forget_send_to_agent,
 )
@@ -70,18 +70,19 @@ def send_message_to_agent_async(self: "Agent", message: str, other_agent_id: str
     return "Successfully sent message"
 
 
-def send_message_to_agents_matching_all_tags(self: "Agent", message: str, tags: List[str]) -> List[str]:
+def send_message_to_agents_matching_tags(self: "Agent", message: str, match_all: List[str], match_some: List[str]) -> List[str]:
     """
-    Sends a message to all agents within the same organization that match all of the specified tags. Messages are dispatched in parallel for improved performance, with retries to handle transient issues and timeouts to ensure responsiveness. This function enforces a limit of 100 agents and does not support pagination (cursor-based queries). Each agent must match all specified tags (`match_all_tags=True`) to be included.
+    Sends a message to all agents within the same organization that match the specified tag criteria. Agents must possess *all* of the tags in `match_all` and *at least one* of the tags in `match_some` to receive the message.
 
     Args:
         message (str): The content of the message to be sent to each matching agent.
-        tags (List[str]): A list of tags that an agent must possess to receive the message.
+        match_all (List[str]): A list of tags that an agent must possess to receive the message.
+        match_some (List[str]): A list of tags where an agent must have at least one to qualify.
 
     Returns:
-        List[str]: A list of responses from the agents that matched all tags. Each
-        response corresponds to a single agent. Agents that do not respond will not
-        have an entry in the returned list.
+        List[str]: A list of responses from the agents that matched the filtering criteria. Each
+        response corresponds to a single agent. Agents that do not respond will not have an entry
+        in the returned list.
     """
 
-    return asyncio.run(_send_message_to_agents_matching_all_tags_async(self, message, tags))
+    return asyncio.run(_send_message_to_agents_matching_tags_async(self, message, match_all, match_some))

{letta_nightly-0.6.37.dev20250310103931 → letta_nightly-0.6.38.dev20250312104155}/letta/functions/helpers.py

@@ -518,8 +518,16 @@ def fire_and_forget_send_to_agent(
         run_in_background_thread(background_task())
 
 
-async def _send_message_to_agents_matching_all_tags_async(sender_agent: "Agent", message: str, tags: List[str]) -> List[str]:
-    log_telemetry(sender_agent.logger, "_send_message_to_agents_matching_all_tags_async start", message=message, tags=tags)
+async def _send_message_to_agents_matching_tags_async(
+    sender_agent: "Agent", message: str, match_all: List[str], match_some: List[str]
+) -> List[str]:
+    log_telemetry(
+        sender_agent.logger,
+        "_send_message_to_agents_matching_tags_async start",
+        message=message,
+        match_all=match_all,
+        match_some=match_some,
+    )
     server = get_letta_server()
 
     augmented_message = (
@@ -529,9 +537,22 @@ async def _send_message_to_agents_matching_all_tags_async(sender_agent: "Agent",
     )
 
     # Retrieve up to 100 matching agents
-    log_telemetry(sender_agent.logger, "_send_message_to_agents_matching_all_tags_async listing agents start", message=message, tags=tags)
-    matching_agents = server.agent_manager.list_agents(actor=sender_agent.user, tags=tags, match_all_tags=True, limit=100)
-    log_telemetry(sender_agent.logger, "_send_message_to_agents_matching_all_tags_async  listing agents finish", message=message, tags=tags)
+    log_telemetry(
+        sender_agent.logger,
+        "_send_message_to_agents_matching_tags_async listing agents start",
+        message=message,
+        match_all=match_all,
+        match_some=match_some,
+    )
+    matching_agents = server.agent_manager.list_agents_matching_tags(actor=sender_agent.user, match_all=match_all, match_some=match_some)
+
+    log_telemetry(
+        sender_agent.logger,
+        "_send_message_to_agents_matching_tags_async  listing agents finish",
+        message=message,
+        match_all=match_all,
+        match_some=match_some,
+    )
 
     # Create a system message
     messages = [MessageCreate(role=MessageRole.system, content=augmented_message, name=sender_agent.agent_state.name)]
@@ -559,7 +580,13 @@ async def _send_message_to_agents_matching_all_tags_async(sender_agent: "Agent",
         else:
             final.append(r)
 
-    log_telemetry(sender_agent.logger, "_send_message_to_agents_matching_all_tags_async finish", message=message, tags=tags)
+    log_telemetry(
+        sender_agent.logger,
+        "_send_message_to_agents_matching_tags_async finish",
+        message=message,
+        match_all=match_all,
+        match_some=match_some,
+    )
     return final
 
 

{letta_nightly-0.6.37.dev20250310103931 → letta_nightly-0.6.38.dev20250312104155}/letta/llm_api/anthropic.py

@@ -53,6 +53,11 @@ MODEL_LIST = [
         "name": "claude-3-opus-20240229",
         "context_window": 200000,
     },
+    # latest
+    {
+        "name": "claude-3-opus-latest",
+        "context_window": 200000,
+    },
     ## Sonnet
     # 3.0
     {
@@ -69,11 +74,21 @@ MODEL_LIST = [
         "name": "claude-3-5-sonnet-20241022",
         "context_window": 200000,
     },
+    # 3.5 latest
+    {
+        "name": "claude-3-5-sonnet-latest",
+        "context_window": 200000,
+    },
     # 3.7
     {
         "name": "claude-3-7-sonnet-20250219",
         "context_window": 200000,
     },
+    # 3.7 latest
+    {
+        "name": "claude-3-7-sonnet-latest",
+        "context_window": 200000,
+    },
     ## Haiku
     # 3.0
     {
@@ -85,6 +100,11 @@ MODEL_LIST = [
         "name": "claude-3-5-haiku-20241022",
         "context_window": 200000,
     },
+    # 3.5 latest
+    {
+        "name": "claude-3-5-haiku-latest",
+        "context_window": 200000,
+    },
 ]
 
 DUMMY_FIRST_USER_MESSAGE = "User initializing bootup sequence."