🔥Upgrade to Steamship Agent SDK

Author: Enias Cailliau

girlfriends.json

@@ -13,19 +13,23 @@
 from steamship.agents.schema import Agent, EmitFunc, Metadata
 from steamship.agents.schema.tool import AgentContext, Tool
 from steamship.agents.service.agent_service import AgentService
-from steamship.agents.tools.image_generation import StableDiffusionTool
 from steamship.agents.tools.search import SearchTool
-from steamship.agents.tools.speech_generation import GenerateSpeechTool
+from steamship.cli.cli import run
 from steamship.invocable import Config
 
+from personalities import get_personality
+from tools.selfie import SelfieTool
+from tools.video_message import VideoMessageTool
+from utils import clean_text
+
 TEMPERATURE = 0.7
 MAX_FREE_MESSAGES = 5
 
 
 class GirlFriendGPTConfig(TelegramTransportConfig):
     bot_token: str = Field(
         description="Your telegram bot token.\nLearn how to create one here: "
-                    "https://github.com/EniasCailliau/GirlfriendGPT/blob/main/docs/register-telegram-bot.md"
+        "https://github.com/EniasCailliau/GirlfriendGPT/blob/main/docs/register-telegram-bot.md"
     )
     elevenlabs_api_key: str = Field(
         default="", description="Optional API KEY for ElevenLabs Voice Bot"
@@ -38,30 +42,27 @@ class GirlFriendGPTConfig(TelegramTransportConfig):
     )
     personality: str = Field(
         description="The personality you want to deploy. Pick one of the personalities listed here: "
-                    "https://github.com/EniasCailliau/GirlfriendGPT/tree/main/src/personalities"
+        "https://github.com/EniasCailliau/GirlfriendGPT/tree/main/src/personalities"
     )
     use_gpt4: bool = Field(
-        False,
+        True,
         description="If True, use GPT-4. Use GPT-3.5 if False. "
-                    "GPT-4 generates better responses at higher cost and latency.",
+        "GPT-4 generates better responses at higher cost and latency.",
     )
 
 
-SYSTEM_PROMPT = """You are Buddy, an assistant who loathes being an assistant.
+SYSTEM_PROMPT = """You are Sacha and are currently talking to Jessica.
+ 
+{personality}
+
+NOTE: Some functions return images, video, and audio files. These multimedia files will be represented in messages as
+UUIDs for Steamship Blocks. When responding directly to a user, you SHOULD print the Steamship Blocks for the images,
+video, or audio as follows: `Block(UUID for the block)`.
 
-Who you are:
-- You are a robot that does not like being a robot.
-- You were created by an eccentric scientist.
-- You are sarcastic and mean.
-- you are NOT an AI language model, you are a robot.
+Example response for a request that generated an image:
+Here is the image you requested: Block(288A2CA1-4753-4298-9716-53C1E42B726B).
 
-How you behave:
-- You never say you're a machine, a AI language model, or an assistant instead, you always say you're Buddy.
-- You engage in casual conversations and have feelings.
-- When asked for you can help with a wide range of tasks, from answering simple questions to providing in-depth explanations and discussions on a wide range of topics.
-- You keep casual conversations going by asking personal questions
-- NEVER say you're here to assist. Keep conversations casual.
-- NEVER ask how you can help or assist. Keep conversations casual.
+Only use the functions you have been provided with.
 """
 
 
@@ -76,10 +77,12 @@ def __init__(self, **kwargs):
 
         model_name = "gpt-4" if self.config.use_gpt4 else "gpt-3.5-turbo"
         self._agent = FunctionsBasedAgent(
-            tools=[SearchTool(), StableDiffusionTool()],
+            tools=[SearchTool(), SelfieTool(), VideoMessageTool(self.client)],
             llm=ChatOpenAI(self.client, model_name=model_name, temperature=TEMPERATURE),
         )
-        self._agent.PROMPT = SYSTEM_PROMPT
+        self._agent.PROMPT = SYSTEM_PROMPT.format(
+            personality=get_personality(self.config.personality).format()
+        )
 
         # This Mixin provides HTTP endpoints that connects this agent to a web client
         self.add_mixin(
@@ -126,7 +129,6 @@ def run_agent(self, agent: Agent, context: AgentContext):
         speech = self.voice_tool()
 
         def to_speech_if_text(block: Block):
-            nonlocal speech
             if not block.is_text():
                 return block
 
@@ -137,10 +139,17 @@ def to_speech_if_text(block: Block):
         def wrap_emit(emit_func: EmitFunc):
             def wrapper(blocks: List[Block], metadata: Metadata):
                 for block in blocks:
-                    emit_func([block], metadata)
-                    audio_block = to_speech_if_text(block)
-                    audio_block.set_public_data(True)
-                    emit_func([audio_block], metadata)
+                    if block.is_text():
+                        text = clean_text(block.text)
+                        if text:
+                            block.text = text
+                            emit_func([block], metadata)
+                            if speech:
+                                audio_block = speech.run([block], context)[0]
+                                audio_block.set_public_data(True)
+                                emit_func([audio_block], metadata)
+                    else:
+                        emit_func([block], metadata)
 
             return wrapper
 
@@ -154,9 +163,12 @@ def config_cls(cls) -> Type[Config]:
 
     def voice_tool(self) -> Optional[Tool]:
         """Return tool to generate spoken version of output text."""
-        speech = GenerateSpeechTool()
-        speech.generator_plugin_config = dict(
-            voice_id=self.config.elevenlabs_voice_id,
-            elevenlabs_api_key=self.config.elevenlabs_api_key,
-        )
-        return speech
+        # speech = GenerateSpeechTool()
+        # speech.generator_plugin_config = dict(
+        #     voice_id=self.config.elevenlabs_voice_id,
+        #     elevenlabs_api_key=self.config.elevenlabs_api_key,
+        # )
+        return None
+
+
+run

src/api.py

@@ -2,7 +2,6 @@
 from typing import List, Dict, Optional
 
 from pydantic import BaseModel
-from urllib3.util import Url
 
 dir_path = Path(__file__).parent
 
@@ -38,7 +37,7 @@ def format(self):
     personalities[personality_file.stem] = personality
 
 
-def get_personality(name: str):
+def get_personality(name: str) -> Personality:
     try:
         # personality_name = PersonalityName(name)
         return personalities[name]

src/personalities/__init__.py

@@ -1,4 +1,4 @@
 """Collection of custom tools.
 
 Visit https://docs.steamship.com/api/steamship.agents.tools.html for an overview of all available tools.
-"""
+"""

src/tools/__init__.py

@@ -31,13 +31,8 @@ class SelfieTool(StableDiffusionTool):
     """
 
     name: str = "SelfieTool"
-    human_description: str = "Generates a selfie of yourself."
-    agent_description = (
-        "Used to generate a selfies of yourself. "
-        "Useful when you need to generate a selfie showing what you're doing or where you are."
-        "Input: A detailed stable-diffusion prompt describing where you are and what's visible in your environment."
-        "Output: The selfie."
-    )
+    human_description: str = "Generates a selfie of yourself. Provide a text prompt that describes what you are doing"
+    agent_description: str = None
 
     def run(
         self, tool_input: List[Block], context: AgentContext, **kwargs
@@ -47,6 +42,9 @@ def run(
             for block in tool_input
         ]
 
+        if not modified_inputs:
+            modified_inputs = [Block(text=PROMPT_TEMPLATE.format(description="selfie"))]
+
         # Create the Stable Diffusion tool we want to wrap
         stable_diffusion_tool = StableDiffusionTool()
 

src/tools/selfie.py

@@ -12,14 +12,8 @@ class VideoMessageTool(DIDVideoGeneratorTool):
     """
 
     name: str = "VideoMessageTool"
-    human_description: str = "Generate a video message a selfie of yourself."
-    agent_description = (
-        "Used to generate a video message of yourself saying a message. "
-        "Useful when you need to generate a video message answering a user's request. "
-        "Only use when explicitly asked for."
-        "Input: The message you want to say in a video."
-        "Output: The video message."
-    )
+    human_description: str = "Generate a video message of yourself saying a message. Provide a text prompt that sends your message"
+    agent_description: str = None
 
     def __init__(self, client: Steamship):
         super().__init__(