pavanjava · pavanjava · Oct 4, 2025 · Oct 4, 2025
diff --git a/devops_agent/cli.py b/devops_agent/cli.py
@@ -2,7 +2,7 @@
 from rich.console import Console
 from rich.panel import Panel
 from rich.prompt import Prompt
-from devops_agent.core.devops_agent import execute_devops_agent
+from devops_agent.core.master_agent import execute_master_agent

 console = Console()

@@ -90,7 +90,7 @@ def run_interactive_mode(provider: str, output: str = None, format: str = 'text'
            ))

            try:
-                response = execute_devops_agent(provider=provider, user_query=user_input)
+                response = execute_master_agent(provider=provider, user_query=user_input)
                console.print(f"\n[bold cyan]Assistant:[/bold cyan]\n{response}")

                # Save to output file if specified
@@ -118,7 +118,7 @@ def process_query(provider: str, query: str, output: str = None, format: str = '
    ))

    try:
-        response = execute_devops_agent(provider=provider, user_query=query)
+        response = execute_master_agent(provider=provider, user_query=query)
        console.print(f"\n[bold cyan]Assistant:[/bold cyan]\n{response}")

        if output:

diff --git a/devops_agent/core/devops_agent.py b/devops_agent/core/devops_agent.py
@@ -9,24 +9,32 @@
 from agno.vectordb.qdrant import Qdrant
 from agno.knowledge.embedder.fastembed import FastEmbedEmbedder
 from qdrant_client.http.models import VectorParams, Distance
+from rich.console import Console
+from rich.panel import Panel

 from devops_agent.utils.prompt_generator_from_poml import prompt_from_poml
 from qdrant_client.qdrant_client import QdrantClient

 devops_prompt = prompt_from_poml('devops.poml')

-qclient = QdrantClient(url=os.environ.get('QDRANT_URL'), api_key=os.environ.get('QDRANT_API_KEY'))
-if not qclient.collection_exists("devops-memory"):
-    qclient.create_collection(collection_name="devops-memory", vectors_config=VectorParams(size=768, distance=Distance.COSINE))
+# qclient = QdrantClient(url=os.environ.get('QDRANT_URL'), api_key=os.environ.get('QDRANT_API_KEY'))
+# if not qclient.collection_exists("devops-memory"):
+#     qclient.create_collection(collection_name="devops-memory", vectors_config=VectorParams(size=768, distance=Distance.COSINE))
+#
+# vector_db = Qdrant(collection="devops-memory", url=os.environ.get('QDRANT_URL'),
+#                    api_key=os.environ.get('QDRANT_API_KEY'),
+#                    embedder=FastEmbedEmbedder(id="snowflake/snowflake-arctic-embed-m"))
+#
+# # Create knowledge base
+# knowledge = Knowledge(vector_db=vector_db)

-vector_db = Qdrant(collection="devops-memory", url=os.environ.get('QDRANT_URL'),
-                   api_key=os.environ.get('QDRANT_API_KEY'),
-                   embedder=FastEmbedEmbedder(id="snowflake/snowflake-arctic-embed-m"))
+console = Console()

-# Create knowledge base
-knowledge = Knowledge(vector_db=vector_db)
-
-def execute_devops_agent(provider: str, user_query: str) -> str:
+def execute_devops_agent(provider: str, user_query: str = None) -> Agent:
+    console.print(Panel.fit(
+        "[bold cyan]DevOps Agent Executing...[/bold cyan]",
+        border_style="cyan"
+    ))
    llm_provider = provider.lower().strip()
    if llm_provider == 'openai':
        model = OpenAIChat(id="gpt-5-mini", api_key=os.environ.get('OPENAI_API_KEY'))
@@ -38,21 +46,23 @@ def execute_devops_agent(provider: str, user_query: str) -> str:
        model = OpenAIChat(id="gpt-5-mini"), #default

    devops_assist = Agent(
-        name="DevOps Assist",
+        name="DevOps Agent",
        model=model,
        description="You help answer questions about the devops domain.",
        instructions=devops_prompt,
-        knowledge=knowledge,
-        stream_intermediate_steps=True,
-        add_knowledge_to_context=True,
-        add_datetime_to_context=True,
-        add_session_summary_to_context=True,
+        # knowledge=knowledge,
+        # stream_intermediate_steps=True,
+        # add_knowledge_to_context=True,
+        # add_datetime_to_context=True,
+        # add_session_summary_to_context=True,
        markdown=True,
    )

-    response = devops_assist.run(user_query, stream_intermediate_steps=True, retry=3)
+    # response = devops_assist.run(user_query, stream_intermediate_steps=True, retry=3)
+    #
+    # asyncio.run(
+    #     knowledge.add_content_async(text_content=response.content, metadata={"agent_id": response.agent_id, "session_id": response.session_id})
+    # )
+    # return response.content

-    asyncio.run(
-        knowledge.add_content_async(text_content=response.content, metadata={"agent_id": response.agent_id, "session_id": response.session_id})
-    )
-    return response.content
+    return devops_assist
diff --git a/devops_agent/core/kubernetes_agent.py b/devops_agent/core/kubernetes_agent.py
@@ -0,0 +1,72 @@
+import asyncio
+import os
+
+from agno.agent import Agent
+from agno.knowledge import Knowledge
+from agno.models.openai import OpenAIChat
+from agno.models.anthropic import Claude
+from agno.models.google.gemini import Gemini
+from agno.vectordb.qdrant import Qdrant
+from agno.knowledge.embedder.fastembed import FastEmbedEmbedder
+from qdrant_client.http.models import VectorParams, Distance
+from rich.console import Console
+from rich.panel import Panel
+
+from devops_agent.utils.prompt_generator_from_poml import prompt_from_poml
+from qdrant_client.qdrant_client import QdrantClient
+
+k8s_prompt = prompt_from_poml('kubernetes.poml')
+
+# qclient = QdrantClient(url=os.environ.get('QDRANT_URL'), api_key=os.environ.get('QDRANT_API_KEY'))
+# if not qclient.collection_exists("devops-memory"):
+#     qclient.create_collection(collection_name="devops-memory",
+#                               vectors_config=VectorParams(size=768, distance=Distance.COSINE))
+#
+# vector_db = Qdrant(collection="devops-memory", url=os.environ.get('QDRANT_URL'),
+#                    api_key=os.environ.get('QDRANT_API_KEY'),
+#                    embedder=FastEmbedEmbedder(id="snowflake/snowflake-arctic-embed-m"))
+#
+# # Create knowledge base
+# knowledge = Knowledge(vector_db=vector_db)
+
+console = Console()
+
+def execute_k8s_agent(provider: str, user_query: str = None) -> Agent:
+
+    console.print(Panel.fit(
+        "[bold cyan]Kubernetes Agent Executing...[/bold cyan]",
+        border_style="cyan"
+    ))
+
+    llm_provider = provider.lower().strip()
+    if llm_provider == 'openai':
+        model = OpenAIChat(id="gpt-5-mini", api_key=os.environ.get('OPENAI_API_KEY'))
+    elif llm_provider == 'anthropic':
+        model = Claude(id="claude-sonnet-4-5-20250929", temperature=0.6, api_key=os.environ.get('ANTHROPIC_API_KEY'))
+    elif llm_provider == 'google':
+        model = Gemini(id="gemini-2.5-flash", temperature=0.6, api_key=os.environ.get('GEMINI_API_KEY'))
+    else:
+        model = OpenAIChat(id="gpt-5-mini"),  # default
+
+    k8s_assist = Agent(
+        name="Kubernetes Agent",
+        model=model,
+        description="You help answer questions about the kubernetes domain of any infrastructure like Azure(AKS), AWS(EKS), and GCP(GKS)",
+        instructions=k8s_prompt,
+        # knowledge=knowledge,
+        # stream_intermediate_steps=True,
+        # add_knowledge_to_context=True,
+        # add_datetime_to_context=True,
+        # add_session_summary_to_context=True,
+        markdown=True,
+    )
+
+    # response = k8s_assist.run(user_query, stream_intermediate_steps=True, retry=3)
+    #
+    # asyncio.run(
+    #     knowledge.add_content_async(text_content=response.content,
+    #                                 metadata={"agent_id": response.agent_id, "session_id": response.session_id})
+    # )
+    # return response.content
+
+    return k8s_assist
diff --git a/devops_agent/core/master_agent.py b/devops_agent/core/master_agent.py
@@ -0,0 +1,74 @@
+import asyncio
+import os
+
+from agno.knowledge import Knowledge
+from agno.models.openai import OpenAIChat
+from agno.models.anthropic import Claude
+from agno.models.google.gemini import Gemini
+from agno.team import Team
+from agno.vectordb.qdrant import Qdrant
+from agno.knowledge.embedder.fastembed import FastEmbedEmbedder
+from qdrant_client import QdrantClient
+from qdrant_client.http.models import VectorParams, Distance
+
+from devops_agent.core.devops_agent import execute_devops_agent
+from devops_agent.core.kubernetes_agent import execute_k8s_agent
+
+qclient = QdrantClient(url=os.environ.get('QDRANT_URL'), api_key=os.environ.get('QDRANT_API_KEY'))
+if not qclient.collection_exists("devops-memory"):
+    qclient.create_collection(collection_name="devops-memory",
+                              vectors_config=VectorParams(size=768, distance=Distance.COSINE))
+
+vector_db = Qdrant(collection="devops-memory", url=os.environ.get('QDRANT_URL'),
+                   api_key=os.environ.get('QDRANT_API_KEY'),
+                   embedder=FastEmbedEmbedder(id="snowflake/snowflake-arctic-embed-m"))
+
+# Create knowledge base
+knowledge = Knowledge(vector_db=vector_db)
+
+
+def execute_master_agent(provider: str, user_query: str) -> str:
+    llm_provider = provider.lower().strip()
+    if llm_provider == 'openai':
+        model = OpenAIChat(id="gpt-5-mini", api_key=os.environ.get('OPENAI_API_KEY'))
+    elif llm_provider == 'anthropic':
+        model = Claude(id="claude-sonnet-4-5-20250929", temperature=0.6, api_key=os.environ.get('ANTHROPIC_API_KEY'))
+    elif llm_provider == 'google':
+        model = Gemini(id="gemini-2.5-flash", temperature=0.6, api_key=os.environ.get('GEMINI_API_KEY'))
+    else:
+        model = OpenAIChat(id="gpt-5-mini"),  # default
+
+    devops_team = Team(
+        name="Multi Cloud and Devops Team",
+        model=model,
+        respond_directly=True,
+        members=[
+            execute_devops_agent(provider=provider),
+            execute_k8s_agent(provider=provider),
+        ],
+        markdown=True,
+        instructions=[
+            "You are a intelligent router that directs questions to the appropriate agent.",
+            "If the user asks in a non devops or k8s question whose agent is not a team member, respond in English with:",
+            "'I can only answer in the following technologies: Devops & Kubernetes Architecture on Multiple clouds. Please ask your question in one of these technologies.'",
+            "Always check the technology or domain of the user's input before routing to an agent.",
+            "For unsupported technologies like coding, flowcharts, analytics etc respond in English with the above message.",
+        ],
+        knowledge=knowledge,
+        determine_input_for_members=True,
+        stream_intermediate_steps=True,
+        add_knowledge_to_context=True,
+        add_datetime_to_context=True,
+        add_session_summary_to_context=True,
+        show_members_responses=True,
+    )
+
+    response = devops_team.run(user_query, stream_intermediate_steps=True, retry=3)
+
+    # saved the response to knowledge in async mode
+    asyncio.run(
+        knowledge.add_content_async(text_content=f"question: {user_query}, Assistant: {response.content}",
+                                    metadata={"agent_id": response.team_id, "session_id": response.session_id})
+    )
+
+    return response.content