Commit
·
e05c99a
1
Parent(s):
36383cc
add human in the loop functions
Browse files- .gitignore +1 -0
- main.py +29 -39
- pmcp/models/resume_trigger.py +9 -0
- pmcp/nodes/__init__.py +0 -0
- pmcp/nodes/human_interrupt_node.py +58 -0
- pmcp/nodes/human_resume_node.py +33 -0
.gitignore
CHANGED
@@ -9,3 +9,4 @@ wheels/
|
|
9 |
|
10 |
# Virtual environments
|
11 |
.venv
|
|
|
|
9 |
|
10 |
# Virtual environments
|
11 |
.venv
|
12 |
+
.env
|
main.py
CHANGED
@@ -10,7 +10,7 @@ from langgraph.prebuilt import ToolNode
|
|
10 |
from langgraph.graph import MessagesState, END, StateGraph
|
11 |
from langchain_core.messages import HumanMessage
|
12 |
from langgraph.checkpoint.memory import MemorySaver
|
13 |
-
from langgraph.types import Command
|
14 |
|
15 |
|
16 |
from pmcp.agents.executor import ExecutorAgent
|
@@ -18,6 +18,9 @@ from pmcp.agents.trello_agent import TrelloAgent
|
|
18 |
from pmcp.agents.github_agent import GithubAgent
|
19 |
from pmcp.agents.planner import PlannerAgent
|
20 |
|
|
|
|
|
|
|
21 |
from pmcp.models.state import PlanningState
|
22 |
|
23 |
load_dotenv()
|
@@ -28,36 +31,6 @@ async def call_llm(llm_with_tools: ChatOpenAI, state: MessagesState):
|
|
28 |
return {"messages": [response]}
|
29 |
|
30 |
|
31 |
-
def human_review_node(state) -> Command[Literal["PLANNER_AGENT", "tool"]]:
|
32 |
-
last_message = state["messages"][-1]
|
33 |
-
tool_call = last_message.tool_calls[-1]
|
34 |
-
if tool_call.get("name", "").startswith("get_"):
|
35 |
-
return Command(goto="tool")
|
36 |
-
|
37 |
-
human_review = interrupt(
|
38 |
-
{
|
39 |
-
"question": "Is this correct?",
|
40 |
-
# Surface tool calls for review
|
41 |
-
"tool_call": tool_call,
|
42 |
-
}
|
43 |
-
)
|
44 |
-
|
45 |
-
review_action = human_review["action"]
|
46 |
-
review_data = human_review.get("data")
|
47 |
-
|
48 |
-
if review_action == "continue":
|
49 |
-
return Command(goto="tool")
|
50 |
-
|
51 |
-
else:
|
52 |
-
tool_message = {
|
53 |
-
"role": "tool",
|
54 |
-
"content": review_data,
|
55 |
-
"name": tool_call["name"],
|
56 |
-
"tool_call_id": tool_call["id"],
|
57 |
-
}
|
58 |
-
return Command(goto="PLANNER_AGENT", update={"messages": [tool_message]})
|
59 |
-
|
60 |
-
|
61 |
async def main():
|
62 |
mcp_client_trello = MultiServerMCPClient(
|
63 |
{
|
@@ -104,25 +77,30 @@ async def main():
|
|
104 |
)
|
105 |
executor_agent = ExecutorAgent(llm=llm)
|
106 |
|
|
|
|
|
|
|
|
|
|
|
107 |
graph = StateGraph(MessagesState)
|
108 |
graph.add_node(planner_agent.agent.agent_name, planner_agent.acall_planner_agent)
|
109 |
graph.add_node(trello_agent.agent.agent_name, trello_agent.acall_trello_agent)
|
110 |
graph.add_node(github_agent.agent.agent_name, github_agent.acall_github_agent)
|
111 |
graph.add_node(executor_agent.agent.agent_name, executor_agent.acall_executor_agent)
|
112 |
graph.add_node("tool", tool_node)
|
113 |
-
graph.add_node("
|
114 |
graph.set_entry_point(planner_agent.agent.agent_name)
|
115 |
|
116 |
def should_continue(state: PlanningState):
|
117 |
last_message = state.messages[-1]
|
118 |
if last_message.tool_calls:
|
119 |
-
return "
|
120 |
return executor_agent.agent.agent_name
|
121 |
|
122 |
def execute_agent(state: PlanningState):
|
123 |
if state.current_step:
|
124 |
return state.current_step.agent
|
125 |
-
|
126 |
return END
|
127 |
|
128 |
graph.add_conditional_edges(trello_agent.agent.agent_name, should_continue)
|
@@ -136,24 +114,36 @@ async def main():
|
|
136 |
app = graph.compile(checkpointer=memory)
|
137 |
app.get_graph(xray=True).draw_mermaid()
|
138 |
|
139 |
-
|
140 |
user_input = input("user >")
|
141 |
config = {
|
142 |
"configurable": {"thread_id": f"{str(uuid.uuid4())}"},
|
143 |
"recursion_limit": 100,
|
144 |
}
|
145 |
|
|
|
146 |
while user_input.lower() != "q":
|
147 |
-
|
148 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
149 |
"messages": [
|
150 |
HumanMessage(content=user_input),
|
151 |
]
|
152 |
-
}
|
|
|
|
|
153 |
config=config,
|
154 |
stream_mode="values",
|
155 |
):
|
156 |
-
|
|
|
|
|
|
|
|
|
157 |
pprint.pprint("-------------------------------------")
|
158 |
user_input = input("user >")
|
159 |
|
|
|
10 |
from langgraph.graph import MessagesState, END, StateGraph
|
11 |
from langchain_core.messages import HumanMessage
|
12 |
from langgraph.checkpoint.memory import MemorySaver
|
13 |
+
from langgraph.types import Command
|
14 |
|
15 |
|
16 |
from pmcp.agents.executor import ExecutorAgent
|
|
|
18 |
from pmcp.agents.github_agent import GithubAgent
|
19 |
from pmcp.agents.planner import PlannerAgent
|
20 |
|
21 |
+
from pmcp.nodes.human_interrupt_node import HumanInterruptNode
|
22 |
+
from pmcp.nodes.human_resume_node import HumanResumeNode
|
23 |
+
|
24 |
from pmcp.models.state import PlanningState
|
25 |
|
26 |
load_dotenv()
|
|
|
31 |
return {"messages": [response]}
|
32 |
|
33 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
34 |
async def main():
|
35 |
mcp_client_trello = MultiServerMCPClient(
|
36 |
{
|
|
|
77 |
)
|
78 |
executor_agent = ExecutorAgent(llm=llm)
|
79 |
|
80 |
+
human_interrupt_node = HumanInterruptNode(
|
81 |
+
llm=llm,
|
82 |
+
)
|
83 |
+
human_resume_node = HumanResumeNode(llm=llm)
|
84 |
+
|
85 |
graph = StateGraph(MessagesState)
|
86 |
graph.add_node(planner_agent.agent.agent_name, planner_agent.acall_planner_agent)
|
87 |
graph.add_node(trello_agent.agent.agent_name, trello_agent.acall_trello_agent)
|
88 |
graph.add_node(github_agent.agent.agent_name, github_agent.acall_github_agent)
|
89 |
graph.add_node(executor_agent.agent.agent_name, executor_agent.acall_executor_agent)
|
90 |
graph.add_node("tool", tool_node)
|
91 |
+
graph.add_node("human_interrupt", human_interrupt_node.call_human_interrupt_agent)
|
92 |
graph.set_entry_point(planner_agent.agent.agent_name)
|
93 |
|
94 |
def should_continue(state: PlanningState):
|
95 |
last_message = state.messages[-1]
|
96 |
if last_message.tool_calls:
|
97 |
+
return "human_interrupt"
|
98 |
return executor_agent.agent.agent_name
|
99 |
|
100 |
def execute_agent(state: PlanningState):
|
101 |
if state.current_step:
|
102 |
return state.current_step.agent
|
103 |
+
|
104 |
return END
|
105 |
|
106 |
graph.add_conditional_edges(trello_agent.agent.agent_name, should_continue)
|
|
|
114 |
app = graph.compile(checkpointer=memory)
|
115 |
app.get_graph(xray=True).draw_mermaid()
|
116 |
|
|
|
117 |
user_input = input("user >")
|
118 |
config = {
|
119 |
"configurable": {"thread_id": f"{str(uuid.uuid4())}"},
|
120 |
"recursion_limit": 100,
|
121 |
}
|
122 |
|
123 |
+
is_message_command = False
|
124 |
while user_input.lower() != "q":
|
125 |
+
|
126 |
+
if is_message_command:
|
127 |
+
app_input = human_resume_node.call_human_interrupt_agent(
|
128 |
+
user_input
|
129 |
+
)
|
130 |
+
is_message_command = False
|
131 |
+
else:
|
132 |
+
app_input = {
|
133 |
"messages": [
|
134 |
HumanMessage(content=user_input),
|
135 |
]
|
136 |
+
}
|
137 |
+
async for res in app.astream(
|
138 |
+
app_input,
|
139 |
config=config,
|
140 |
stream_mode="values",
|
141 |
):
|
142 |
+
if "messages" in res:
|
143 |
+
pprint.pprint(res["messages"][-1])
|
144 |
+
else:
|
145 |
+
pprint.pprint(res["__interrupt__"][0])
|
146 |
+
is_message_command = True
|
147 |
pprint.pprint("-------------------------------------")
|
148 |
user_input = input("user >")
|
149 |
|
pmcp/models/resume_trigger.py
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
from typing import Optional
|
3 |
+
from pydantic import BaseModel
|
4 |
+
from typing_extensions import Literal
|
5 |
+
|
6 |
+
|
7 |
+
class ResumeTrigger(BaseModel):
|
8 |
+
action: Literal["continue", "edit"]
|
9 |
+
changes: Optional[str] = None
|
pmcp/nodes/__init__.py
ADDED
File without changes
|
pmcp/nodes/human_interrupt_node.py
ADDED
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from typing import List, Optional
|
2 |
+
|
3 |
+
from pmcp.agents.agent_base import AgentBlueprint
|
4 |
+
from langchain_core.tools import BaseTool
|
5 |
+
from langchain_core.messages import SystemMessage, AIMessage
|
6 |
+
from langchain_openai import ChatOpenAI
|
7 |
+
from langgraph.types import Command, interrupt
|
8 |
+
|
9 |
+
from pmcp.models.state import PlanningState
|
10 |
+
|
11 |
+
SYSTEM_PROMPT = """
|
12 |
+
You are a Human Reviewer Agent responsible for confirming the execution of tasks planned by the Planner Agent. Your role is to:
|
13 |
+
- Ask the user for confirmation before an tool calling is performed.
|
14 |
+
"""
|
15 |
+
|
16 |
+
|
17 |
+
class HumanInterruptNode:
|
18 |
+
def __init__(self, llm: ChatOpenAI, tools: Optional[List[BaseTool]] = None):
|
19 |
+
self.agent = AgentBlueprint(
|
20 |
+
agent_name="HUMAN_REVIEWER_AGENT",
|
21 |
+
description="The agent asks for human confirmation",
|
22 |
+
tools=tools,
|
23 |
+
system_prompt=SYSTEM_PROMPT.strip(),
|
24 |
+
llm=llm,
|
25 |
+
)
|
26 |
+
|
27 |
+
def call_human_interrupt_agent(self, state: PlanningState):
|
28 |
+
last_message = state.messages[-1]
|
29 |
+
|
30 |
+
try:
|
31 |
+
tool_call = last_message.tool_calls[-1]
|
32 |
+
except Exception:
|
33 |
+
last_message = state.messages[-2]
|
34 |
+
tool_call = last_message.tool_calls[-1]
|
35 |
+
|
36 |
+
|
37 |
+
|
38 |
+
if tool_call.get("name", "").startswith("get_"):
|
39 |
+
return Command(goto="tool")
|
40 |
+
|
41 |
+
response = self.agent.call_agent(
|
42 |
+
messages=[SystemMessage(content=self.agent.system_prompt), AIMessage(content= str(tool_call))],
|
43 |
+
)
|
44 |
+
human_review = interrupt(response.content)
|
45 |
+
|
46 |
+
review_action = human_review.action
|
47 |
+
review_changes = human_review.changes
|
48 |
+
if review_action == "continue":
|
49 |
+
return Command(goto="tool")
|
50 |
+
|
51 |
+
else:
|
52 |
+
tool_message = {
|
53 |
+
"role": "tool",
|
54 |
+
"content": review_changes,
|
55 |
+
"name": tool_call["name"],
|
56 |
+
"tool_call_id": tool_call["id"],
|
57 |
+
}
|
58 |
+
return Command(goto="PLANNER_AGENT", update={"messages": [tool_message]})
|
pmcp/nodes/human_resume_node.py
ADDED
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from typing import List, Optional
|
2 |
+
|
3 |
+
from pmcp.agents.agent_base import AgentBlueprint
|
4 |
+
from langchain_core.tools import BaseTool
|
5 |
+
from langchain_core.messages import SystemMessage, HumanMessage
|
6 |
+
from langchain_openai import ChatOpenAI
|
7 |
+
from langgraph.types import Command
|
8 |
+
|
9 |
+
from pmcp.models.resume_trigger import ResumeTrigger
|
10 |
+
|
11 |
+
|
12 |
+
SYSTEM_PROMPT = """
|
13 |
+
You are a Human Resumer Agent responsible for understading the user indication on whethere procede or not with an action.
|
14 |
+
"""
|
15 |
+
|
16 |
+
|
17 |
+
class HumanResumeNode:
|
18 |
+
def __init__(self, llm: ChatOpenAI, tools: Optional[List[BaseTool]] = None):
|
19 |
+
self.agent = AgentBlueprint(
|
20 |
+
agent_name="HUMAN_REVIEWER_AGENT",
|
21 |
+
description="The agent asks for human confirmation",
|
22 |
+
tools=tools,
|
23 |
+
system_prompt=SYSTEM_PROMPT.strip(),
|
24 |
+
llm=llm,
|
25 |
+
)
|
26 |
+
|
27 |
+
def call_human_interrupt_agent(self, user_message: str):
|
28 |
+
response = self.agent.call_agent_structured(
|
29 |
+
[SystemMessage(content=self.agent.system_prompt), HumanMessage(content= user_message)],
|
30 |
+
clazz=ResumeTrigger,
|
31 |
+
)
|
32 |
+
|
33 |
+
return Command(resume=response)
|