{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "78fb59cf-89fd-4b30-8a1c-d1ae4bfd3daf",
   "metadata": {},
   "outputs": [],
   "source": [
    "from letta import create_client, Admin\n",
    "from letta.client.client import LocalClient, RESTClient "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "9269eda2-3108-4955-86ab-b406d51f562a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "UUID('00000000-0000-0000-0000-000000000000')"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "client = create_client() \n",
    "client.user_id"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "879710d4-21c7-43ec-8d00-73e618f55693",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ListModelsResponse(models=[LLMConfigModel(model='gpt-4o-mini', model_endpoint_type='openai', model_endpoint='https://api.openai.com/v1', model_wrapper=None, context_window=8192)])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "client.list_models()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "af6ea8eb-fc6b-4de5-ae79-c2b684a81f17",
   "metadata": {},
   "source": [
    "## Create a key from the Admin portal \n",
    "(This is to allow viewing agents on the dev portal) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "715fa669-3fc6-4579-96a9-c4a730f43e29",
   "metadata": {},
   "outputs": [],
   "source": [
    "admin_client = Admin(base_url=\"http://localhost:8283\", token=\"lettaadmin\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "1782934f-7884-4ee7-ad09-5ae33efa3b2e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CreateAPIKeyResponse(api_key='sk-45cc3e1fd35a3fac3a2ad959fc23877a0476181e8b0a5557')"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admin_client.create_key(user_id=client.user_id, key_name=\"key\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "b29bac8d-2a15-45de-a60d-6d94275443f5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Letta.letta.server.server - INFO - Created new agent from config: <letta.agent.Agent object at 0x14e542600>\n"
     ]
    }
   ],
   "source": [
    "agent_state = client.create_agent()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5fbc43c8-9536-4107-a64d-6e702083242b",
   "metadata": {},
   "source": [
    "## Create an agent "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "f0a388b5-2d00-4f3e-8a5b-b768da02ac8e",
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_resume(self, name: str): \n",
    "    \"\"\"\n",
    "    Read the resume data for a candidate given the name\n",
    "\n",
    "    Args: \n",
    "        name (str): Candidate name \n",
    "\n",
    "    Returns: \n",
    "        resume_data (str): Candidate's resume data \n",
    "    \"\"\"\n",
    "    import os\n",
    "    filepath = os.path.join(\"data\", \"resumes\", name.lower().replace(\" \", \"_\") + \".txt\")\n",
    "    #print(\"read\", filepath)\n",
    "    return open(filepath).read()\n",
    "\n",
    "def submit_candidate_for_outreach(self, candidate_name: str, resume: str, justification: str): \n",
    "    \"\"\"\n",
    "    Submit a candidate for outreach. \n",
    "\n",
    "    Args: \n",
    "        candidate_name (str): The name of the candidate\n",
    "        resume (str): The text representation of the candidate's resume \n",
    "        justification (str): Summary reason for why the candidate is good and should be reached out to\n",
    "    \"\"\"\n",
    "    from letta import create_client \n",
    "    client = create_client()\n",
    "    message = \"Reach out to the following candidate. \" \\\n",
    "    + f\"Name: {candidate_name}\\n\" \\\n",
    "    + f\"Resume Data: {resume}\\n\" \\\n",
    "    + f\"Justification: {justification}\"\n",
    "    # NOTE: we will define this agent later \n",
    "    #print(\"submit for outreach\", message)\n",
    "    response = client.send_message(agent_name=\"outreach_agent\", role=\"user\", message=message) # TODO: implement this\n",
    "    #print(respose.messages)\n",
    "\n",
    "# TODO: add an archival andidate tool (provide justification) \n",
    "\n",
    "read_resume_tool = client.create_tool(read_resume) \n",
    "submit_candidate_tool = client.create_tool(submit_candidate_for_outreach)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "d2b0f66f-6cc3-471f-b2c7-49f51f5bbb7b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Letta.letta.server.server - INFO - Created new agent from config: <letta.agent.Agent object at 0x14e542600>\n"
     ]
    }
   ],
   "source": [
    "from letta.memory import ChatMemory\n",
    "\n",
    "company_description = \"The company is called AgentOS and is building AI tools to make it easier to create and deploy LLM agents.\"\n",
    "skills = \"Front-end (React, Typescript), software engineering (ideally Python), and experience with LLMs.\"\n",
    "\n",
    "\n",
    "leadgen_agent = client.create_agent(\n",
    "    name=\"leadgen_agent\", \n",
    "    memory=ChatMemory(\n",
    "        persona=f\"You are responsible to finding good recruiting candidates, for the company description: {company_description}. \" \\\n",
    "        + f\"Ideal canddiates have skills: {skills}. \" \\\n",
    "        + \"Search for candidates by calling the `search_candidates_db` function. \" \\\n",
    "        + \"When you find a good candidate, submit the candidate for outreach with the `submit_candidate_for_outreach` tool. \" \\\n",
    "        + \"Continue to search through the database until there are no more entries. \",\n",
    "        human=\"\",\n",
    "    ), \n",
    "    tools=[read_resume_tool.name, submit_candidate_tool.name]\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f489784-dbc9-4c93-9181-457460b05401",
   "metadata": {},
   "source": [
    "## Cleanup "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "f93c330b-909a-4180-bf6b-166b951977a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "agents = client.list_agents()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "523a382d-f514-46cb-a902-84ee74706f01",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Deleted FierceNucleus\n",
      "Deleted LuxuriousRaccoon\n"
     ]
    }
   ],
   "source": [
    "for agent in agents: \n",
    "    client.delete_agent(agent.id)\n",
    "    print(\"Deleted\", agent.name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e7f1a012-0080-4e68-b26f-7d139a37bad0",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "letta",
   "language": "python",
   "name": "letta"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}