Fix/tool

example/AutoAgent.ipynb

@@ -12,13 +12,13 @@
     "\n",
     "base_url= \"\" # use your own base_url from api provider or local provider like ollama.\n",
     "api_key = \"\" # use your own api_key.\n",
-    "provider = \"openai\" # openai for openai like provider (vLLM or openrouter) and ollama for local use.\n",
+    "# provider = \"openai\" # Deprecated No need for provider any more\n",
     "model = \"gpt-4o-mini\""
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": null,
    "id": "6057cd04",
    "metadata": {},
    "outputs": [
@@ -46,7 +46,7 @@
    "source": [
     "from iragent.tools import get_time_now, simple_termination\n",
     "\n",
-    "factory = AgentFactory(base_url,api_key, model, provider)\n",
+    "factory = AgentFactory(base_url,api_key, model)\n",
     "\n",
     "agent1 = factory.create_agent(name=\"time_reader\",\n",
     "                            system_prompt=\"You are that one who can read time. there is a fucntion named get_time_now(), you can call it whether user ask about time or date.\",\n",

example/BaseMemory.ipynb

@@ -12,7 +12,7 @@
     "\n",
     "base_url= \"\" # use your own base_url from api provider or local provider like ollama.\n",
     "api_key = \"\" # use your own api_key.\n",
-    "provider = \"openai\" # openai for openai like provider (vLLM or openrouter) and ollama for local use.\n",
+    "# provider = \"openai\" # Deprecated No need for provider any more\n",
     "model = \"gpt-4o-mini\"\n"
    ]
   },
@@ -59,7 +59,7 @@
     "from iragent.memory import BaseMemory\n",
     "from iragent.tools import get_time_now, simple_termination\n",
     "\n",
-    "factory = AgentFactory(base_url,api_key, model, provider)\n",
+    "factory = AgentFactory(base_url,api_key, model)\n",
     "\n",
     "agent1 = factory.create_agent(name=\"time_reader\",\n",
     "                            system_prompt=\"You are that one who can read time. there is a fucntion named get_time_now(), you can call it whether user ask about time or date.\",\n",

example/InternetSearch.ipynb

@@ -11,13 +11,13 @@
     "\n",
     "base_url= \"\" # use your own base_url from api provider or local provider like ollama.\n",
     "api_key = \"\" # use your own api_key.\n",
-    "provider = \"openai\" # openai for openai like provider (vLLM or openrouter) and ollama for local use.\n",
+    "# provider = \"openai\" # Deprecated No need for provider any more\n",
     "model = \"gpt-4o-mini\"\n"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": null,
    "id": "10412d76",
    "metadata": {},
    "outputs": [],
@@ -29,7 +29,6 @@
     "    api_key=api_key,\n",
     "    temperature=0.1,\n",
     "    max_token=512,\n",
-    "    provider=\"openai\" # or ollama for local use \n",
     ")\n"
    ]
   },

example/SmartAgentBuilder.ipynb

@@ -19,13 +19,13 @@
    "source": [
     "base_url= \"\" # use your own base_url from api provider or local provider like ollama.\n",
     "api_key = \"\" # use your own api_key.\n",
-    "provider = \"openai\" # openai for openai like provider (vLLM or openrouter) and ollama for local use.\n",
+    "# provider = \"openai\" # Deprecated No need for provider any more\n",
     "model = \"gpt-4o-mini\""
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": null,
    "id": "10561d55",
    "metadata": {},
    "outputs": [
@@ -47,7 +47,6 @@
     "    base_url=base_url,\n",
     "    api_key=api_key,\n",
     "    model=model,\n",
-    "    provider=\"openai\"\n",
     ")\n",
     "\n",
     "sab = SmartAgentBuilder(\n",

example/SmartPrompt.ipynb

@@ -9,13 +9,13 @@
    "source": [
     "base_url= \"\" # use your own base_url from api provider or local provider like ollama.\n",
     "api_key = \"\" # use your own api_key.\n",
-    "provider = \"openai\" # openai for openai like provider (vLLM or openrouter) and ollama for local use.\n",
+    "# provider = \"openai\" # Deprecated No need for provider any more\n",
     "model = \"gpt-4o-mini\""
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": null,
    "id": "686c8d4a",
    "metadata": {},
    "outputs": [
@@ -89,7 +89,6 @@
     "    base_url=base_url,\n",
     "    api_key=api_key,\n",
     "    model=model,\n",
-    "    provider=\"openai\"\n",
     ")\n",
     "\n",
     "sp = SmartPrompt(agent_factory=agent_factory)\n",

example/StructuredOutput.ipynb

@@ -12,7 +12,7 @@
     "\n",
     "base_url= \"\" # use your own base_url from api provider or local provider like ollama.\n",
     "api_key = \"\" # use your own api_key.\n",
-    "provider = \"openai\" # openai for openai like provider (vLLM or openrouter) and ollama for local use.\n",
+    "# provider = \"openai\" # Deprecated No need for provider any more.\n",
     "model = \"gpt-4o-mini\""
    ]
   },
@@ -39,7 +39,6 @@
     "    base_url=base_url,\n",
     "    api_key=api_key,\n",
     "    model=model,\n",
-    "    provider=provider\n",
     ")\n",
     "\n",
     "PROMPT = \"\"\"\n",

example/SummarizerMemory.ipynb

@@ -17,7 +17,7 @@
    "source": [
     "base_url= \"\" # use your own base_url from api provider or local provider like ollama.\n",
     "api_key = \"\" # use your own api_key.\n",
-    "provider = \"openai\" # openai for openai like provider (vLLM or openrouter) and ollama for local use.\n",
+    "# provider = \"openai\" # Deprecated No need for provider any more\n",
     "model = \"gpt-4o-mini\""
    ]
   },
@@ -31,7 +31,7 @@
     "from iragent.agent import AgentFactory\n",
     "from iragent.memory import SummarizerMemory\n",
     "\n",
-    "factory = AgentFactory(base_url, api_key, model, provider)\n",
+    "factory = AgentFactory(base_url, api_key, model)\n",
     "\n",
     "def log_task(task: str) -> str:\n",
     "    return f\"Task '{task}' logged successfully.\"\n",

iragent/agent.py

@@ -1,6 +1,7 @@
 import inspect
 import json
 import re
+import warnings
 from typing import Any, Callable, Dict, List, get_type_hints
 
 import requests
@@ -79,13 +80,26 @@ def call_message(self, message: Message, **kwargs) -> str:
         # Add to memory if it is first time
         if self.memory:
             self.memory.add_history(user_msg)
+        # Provider will be removed in v0.1.8
+        # TODO: Remove provider and use just one of them.
+        if self.provider:
+            warnings.warn(
+                "'provider' is deprecated and will be removed in the next release. "
+                "You no longer need to define 'provider' — the package will automatically "
+                "select the appropriate backend.",
+                FutureWarning,  # or DeprecationWarning if you want it hidden by default
+                stacklevel=2
+            )
 
-        if self.provider == "openai":
-            res =  self._call_openai(msgs=msgs, message=message, **kwargs)
-        elif self.provider == "ollama":
-            res = self._call_ollama_v2(msgs=msgs, message=message)
+            if self.provider == "openai":
+                res =  self._call_openai(msgs=msgs, message=message, **kwargs)
+            elif self.provider == "ollama":
+                res = self._call_ollama_v2(msgs=msgs, message=message)
+            else:
+                raise ValueError(f"Unsupported provider: {self.provider}")
+
         else:
-            raise ValueError(f"Unsupported provider: {self.provider}")
+            res = self._call_ollama_v2(msgs=msgs, message=message)
 
 
         if self.memory:
@@ -183,7 +197,8 @@ def _call_ollama_v2(self, msgs: List[Dict], message: Message) -> Message:
             max_tokens=self.max_token,
             temperature=self.temprature,
         )
-
+        if self.response_format:
+            kwargs["response_format"] = self.response_format
         if self.fn:
             kwargs["tools"] = [{"type": "function", "function": f} for f in self.fn]
         response = self.client.chat.completions.create(**kwargs)
@@ -198,7 +213,7 @@ def _call_ollama_v2(self, msgs: List[Dict], message: Message) -> Message:
                 followup = self.client.chat.completions.create(
                     model=self.model,
                     messages=msgs
-                    + [msg, {"role": "tool", "name": fn_name, "content": str(result)}],
+                    + [msg, {"role": "tool","tool_call_id": msg.tool_calls[0].id, "name": fn_name, "content": str(result)}],
                     max_tokens=self.max_token,
                     temperature=self.temprature,
                 )
@@ -208,6 +223,18 @@ def _call_ollama_v2(self, msgs: List[Dict], message: Message) -> Message:
                     content=followup.choices[0].message.content.strip(),
                     metadata={"reply_to": message.metadata.get("message_id")},
                 )
+        # Handle response format
+        if self.response_format:
+            try:
+                parsed_content = json.loads(msg.content)
+            except json.JSONDecodeError:
+                parsed_content = {"error": "Invalid JSON response", "raw": msg.content}
+            return Message(
+                sender=self.name,
+                reciever=self.next_agent or message.sender,
+                content=parsed_content,
+                metadata={"reply_to": message.metadata.get("message_id")},
+            )
 
         return Message(
             sender=self.name,
@@ -242,7 +269,7 @@ def _call_openai(self, msgs: List[Dict], message: Message, **kwargs) -> Message:
                     messages=msgs
                     + [
                         msg,
-                        {"role": "function", "name": fn_name, "content": str(result)},
+                        {"role": "function", "tool_call_id": msg.tool_calls[0].id, "name": fn_name, "content": str(result)},
                     ],
                     max_tokens=self.max_token,
                     temperature=self.temprature,
@@ -366,7 +393,7 @@ class AgentFactory:
             Creates and returns a new Agent instance using the shared
             configuration and any additional keyword arguments.    
     """
-    def __init__(self, base_url: str, api_key: str, model: str, provider: str = "openai") -> Agent:
+    def __init__(self, base_url: str, api_key: str, model: str, provider: str = None) -> Agent:
         self.base_url = base_url
         self.api_key = api_key
         self.model = model

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "iragent"
-version = "0.1.6"
+version = "0.1.7"
 description = "A simple multi-agent framework"
 authors = [{ name = "Parsa Bakhtiari", email = "spacenavard1@gmail.com" }]
 readme = "README.md"
@@ -13,21 +13,19 @@ dependencies = [
     "beautifulsoup4",
     "lxml",
     "nltk",
-    "pytest"
+    "pytest",
+    "tiktoken",
+    "numpy"
 ]
 # For RAG usage
 [project.optional-dependencies]
 rag = [
     "faiss-cpu",
     "sentence-transformers",
-    "numpy",
-    "tiktoken"
 ]
 rag-gpu = [
     "faiss-gpu",
     "sentence-transformers",
-    "numpy",
-    "tiktoken"
     ]
 
 [build-system]