Merge pull request #158 from restackio/agentBump

Agents example bump 0.0.57

Author: aboutphilippe

agent_chat/README.md

@@ -48,7 +48,7 @@ You can run workflows from the UI by clicking the "Run" button.
 
 You can run workflows from the API by using the generated endpoint:
 
-`POST http://localhost:6233/api/workflows/AgentWorkflow`
+`POST http://localhost:6233/api/agents/AgentChat`
 
 ### from any client
 
@@ -58,7 +58,7 @@ You can run workflows with any client connected to Restack, for example:
 poetry run schedule
 ```
 
-executes `schedule_workflow.py` which will connect to Restack and execute the `AgentWorkflow` workflow.
+executes `schedule_agent.py` which will connect to Restack and execute the `AgentChat` agent.
 
 ## Send events to the Agent
 
@@ -76,7 +76,7 @@ And see the events in the run:
 
 You can send events to the agent by using the following endpoint:
 
-`PUT http://localhost:6233/api/workflows/AgentWorkflow/:workflowId/:runId`
+`PUT http://localhost:6233/api/agents/AgentChat/:agentId/:runId`
 
 with the payload:
 

agent_chat/chat_post.png

@@ -0,0 +1,26 @@
+import asyncio
+from restack_ai import Restack
+async def main(agent_id:str,run_id:str):
+
+    client = Restack()
+
+    await client.send_agent_event(
+        agent_id=agent_id,
+        run_id=run_id,
+        event_name="message",
+        event_input={"content": "Tell me another joke"}
+    )
+
+    await client.send_agent_event(
+        agent_id=agent_id,
+        run_id=run_id,
+        event_name="end",
+    )
+
+    exit(0)
+
+def run_event_agent():
+    asyncio.run(main(agent_id="your-agent-id", run_id="your-run-id"))
+
+if __name__ == "__main__":
+    run_event_agent()

agent_chat/chat_put.png

@@ -13,10 +13,8 @@ packages = [{include = "src"}]
 python = ">=3.10,<4.0"
 watchfiles = "^1.0.0"
 pydantic = "^2.10.4"
-
-# Build system configuration
 openai = "^1.60.2"
-restack-ai = "^0.0.55"
+restack-ai = "^0.0.57"
 
 [build-system]
 requires = ["poetry-core"]
@@ -26,5 +24,5 @@ build-backend = "poetry.core.masonry.api"
 [tool.poetry.scripts]
 dev = "src.services:watch_services"
 services = "src.services:run_services"
-schedule = "schedule_workflow:run_schedule_workflow"
-event = "event_workflow:run_event_workflow"
+schedule = "schedule_agent:run_schedule_agent"
+event = "event_agent:run_event_agent"

agent_chat/chat_run.png

@@ -1,27 +1,27 @@
 import asyncio
 import time
 from restack_ai import Restack
-from src.workflows.workflow import AgentChatInput
+from agent_chat.src.agents.agent import AgentChatInput
 async def main():
 
     client = Restack()
 
     workflow_id = f"{int(time.time() * 1000)}-AgentChat"
-    run_id = await client.schedule_workflow(
-        workflow_name="AgentChat",
-        workflow_id=workflow_id,
+    run_id = await client.schedule_agent(
+        agent_name="AgentChat",
+        agent_id=workflow_id,
         input=AgentChatInput(message="Tell me a joke")
-    )
+    ) 
 
-    await client.get_workflow_result(
-        workflow_id=workflow_id,
+    await client.get_agent_result(
+        agent_id=workflow_id,
         run_id=run_id
     )
 
     exit(0)
 
-def run_schedule_workflow():
+def run_schedule_agent():
     asyncio.run(main())
 
 if __name__ == "__main__":
-    run_schedule_workflow()
+    run_schedule_agent()

agent_chat/event_agent.py

@@ -1,7 +1,7 @@
 from datetime import timedelta
 from typing import List
 from pydantic import BaseModel
-from restack_ai.workflow import workflow, import_functions, log
+from restack_ai.agent import agent, import_functions, log
 
 with import_functions():
     from src.functions.llm_chat import llm_chat, LlmChatInput, Message
@@ -12,26 +12,26 @@ class MessageEvent(BaseModel):
 class EndEvent(BaseModel):
     end: bool
 
-@workflow.defn()
+@agent.defn()
 class AgentChat:
     def __init__(self) -> None:
         self.end = False
         self.messages = []
-    @workflow.event
+    @agent.event
     async def message(self, message: MessageEvent) -> List[Message]:
         log.info(f"Received message: {message.content}")
         self.messages.append({"role": "user", "content": message.content})
-        assistant_message = await workflow.step(llm_chat, LlmChatInput(messages=self.messages), start_to_close_timeout=timedelta(seconds=120))
+        assistant_message = await agent.step(llm_chat, LlmChatInput(messages=self.messages), start_to_close_timeout=timedelta(seconds=120))
         self.messages.append(assistant_message)
         return self.messages
-    @workflow.event
+    @agent.event
     async def end(self, end: EndEvent) -> EndEvent:
         log.info(f"Received end")
         self.end = True
         return end
-    @workflow.run
+    @agent.run
     async def run(self, input: dict):
-        await workflow.condition(
+        await agent.condition(
             lambda: self.end
         )
         return

agent_chat/event_workflow.py

@@ -2,14 +2,14 @@
 import os
 from src.functions.llm_chat import llm_chat
 from src.client import client
-from src.workflows.workflow import AgentChat
+from src.agents.agent import AgentChat
 from watchfiles import run_process
 import webbrowser
 
 async def main():
 
     await client.start_service(
-        workflows=[AgentChat],
+        agents=[AgentChat],
         functions=[llm_chat]
     )
 

agent_chat/pyproject.toml

@@ -1,9 +1,9 @@
 # Restack AI - Agent Tool
 
 This repository contains a tool for an AI agent.
-It demonstrates how to set up a basic workflow and functions.
+It demonstrates how to set up a basic agent with functions calling.
 
-Follow [the step-by-step tutorial](https://docs.restack.io/examples/projects/agent#tutorial) on how to customize the agent tool example/ 
+Follow [the step-by-step tutorial](https://docs.restack.io/examples/projects/agent#tutorial) on how to customize the agent tool example/
 
 ## Prerequisites
 
@@ -44,24 +44,27 @@ Duplicate the `env.example` file and rename it to `.env`.
 
 Obtain a Restack API Key to interact with the 'restack-c1' model at no cost from [console.restack.io](https://console.restack.io)
 
-## Run workflows
+## Run agents
 
 ### from UI
 
-You can run workflows from the UI by clicking the "Run" button.
+You can run agents from the UI by clicking the "Run" button.
 
-![Run workflows from UI](./screenshot-endpoints.png)
+![Run agents from UI](./chat_post.png)
 
 ### from API
 
-You can run workflows from the API by using the generated endpoint:
+You can run agents from the API by using the generated endpoint:
 
-`POST http://localhost:6233/api/workflows/AgentChatToolFunctions`
+`POST http://localhost:6233/api/agents/AgentChatToolFunctions`
+
+## Send an event to the agent
+
+### from UI
 
-## Send an event to the 
 ```
 {
-  "workflowId": "{workflow_id}",
+  "agentId": "{agent_id}",
   "runId": "{run_id}",
   "eventName": "message",
   "eventInput": {
@@ -70,6 +73,29 @@ You can run workflows from the API by using the generated endpoint:
 }
 ```
 
+![Send event to agent](./chat_put.png)
+
+You can send events to the agent by using the generated endpoint:
+
+`PUT http://localhost:6233/api/agents/AgentChatToolFunctions/:agentId/:runId`
+
+and the payload:
+
+```
+{
+  "eventName": "message",
+  "eventInput": {
+    "content": "What clothes are currently on sales?"
+  }
+}
+```
+
+## See the agent run
+
+You cna replay and follow the agent run in the UI.
+
+![Replay agent run](./chat_run.png)
+
 Now, you can simply trigger more events from the Developer UI by clicking in the timeline on 'Send again' for the event and change the payload.
 
 ![Send another message from UI](./event-send-again.png)

agent_chat/schedule_workflow.py agent_chat/schedule_agent.py

@@ -17,7 +17,7 @@ requests = "2.32.3"
 
 # Build system configuration
 openai = "^1.60.2"
-restack-ai = "^0.0.55"
+restack-ai = "^0.0.57"
 [build-system]
 requires = ["poetry-core"]
 build-backend = "poetry.core.masonry.api"

agent_chat/src/workflows/__init__.py agent_chat/src/agents/__init__.py

@@ -1,7 +1,7 @@
 from datetime import timedelta
 from typing import List
 from pydantic import BaseModel
-from restack_ai.workflow import workflow, import_functions, log
+from restack_ai.agent import agent, import_functions, log
 
 
 with import_functions():
@@ -17,13 +17,13 @@ class MessageEvent(BaseModel):
 class EndEvent(BaseModel):
     end: bool
 
-@workflow.defn()
+@agent.defn()
 class AgentChatToolFunctions:
     def __init__(self) -> None:
         self.end = False
         self.messages = []
 
-    @workflow.event
+    @agent.event
     async def message(self, message: MessageEvent) -> List[Message]:
         log.info(f"Received message: {message.content}")
 
@@ -45,7 +45,7 @@ async def message(self, message: MessageEvent) -> List[Message]:
         system_content = "You are a helpful assistant that can help with sales data."
 
         self.messages.append(Message(role="user", content=message.content or ""))
-        completion = await workflow.step(llm_chat, LlmChatInput(messages=self.messages, tools=tools, system_content=system_content), start_to_close_timeout=timedelta(seconds=120))
+        completion = await agent.step(llm_chat, LlmChatInput(messages=self.messages, tools=tools, system_content=system_content), start_to_close_timeout=timedelta(seconds=120))
 
         log.info(f"completion: {completion}")
 
@@ -66,10 +66,10 @@ async def message(self, message: MessageEvent) -> List[Message]:
 
                         log.info(f"calling {name} with args: {args}")
 
-                        result = await workflow.step(lookupSales, input=LookupSalesInput(category=args.category), start_to_close_timeout=timedelta(seconds=120))
+                        result = await agent.step(lookupSales, input=LookupSalesInput(category=args.category), start_to_close_timeout=timedelta(seconds=120))
                         self.messages.append(Message(role="tool", tool_call_id=tool_call.id, content=str(result)))
 
-                        completion_with_tool_call = await workflow.step(llm_chat, LlmChatInput(messages=self.messages, system_content=system_content), start_to_close_timeout=timedelta(seconds=120))
+                        completion_with_tool_call = await agent.step(llm_chat, LlmChatInput(messages=self.messages, system_content=system_content), start_to_close_timeout=timedelta(seconds=120))
                         self.messages.append(Message(role="assistant", content=completion_with_tool_call.choices[0].message.content or ""))
 
                     ## Step 4: Add your new function to the match case and append the result to the messages
@@ -90,15 +90,15 @@ async def message(self, message: MessageEvent) -> List[Message]:
 
         return self.messages
 
-    @workflow.event
+    @agent.event
     async def end(self, end: EndEvent) -> EndEvent:
         log.info(f"Received end")
         self.end = True
         return {"end": True}
 
-    @workflow.run
+    @agent.run
     async def run(self, input: dict):
-        await workflow.condition(
+        await agent.condition(
             lambda: self.end
         )
         return

agent_chat/src/workflows/workflow.py agent_chat/src/agents/agent.py

@@ -13,7 +13,7 @@
 async def main():
 
     await client.start_service(
-        workflows=[AgentChatToolFunctions],
+        agents=[AgentChatToolFunctions],
         ## Step 6: Add your new function to the functions list -> functions=[lookupSales, llm_chat, new_function]
         functions=[lookupSales, llm_chat]
     )