š« CAMEL: The first and the best multi-agent framework. Finding the Scaling Law of Agents. https://www.camel-ai.org
Python agentai-societiesartificial-intelligencecommunicative-aicooperative-aideep-learninglarge-language-modelsmulti-agent-systemsnatural-language-processing
# ========= Copyright 2023-2026 @ CAMEL-AI.org. All Rights Reserved. =========
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ========= Copyright 2023-2026 @ CAMEL-AI.org. All Rights Reserved. =========
import os
import streamlit as st
import json
import base64
from typing import List, Optional, Union
from pydantic import BaseModel, Field
from camel.toolkits.pptx_toolkit import PPTXToolkit
from camel.models import ModelFactory
from camel.types import ModelPlatformType, ModelType
from camel.agents import ChatAgent
# --- Pydantic Models for Structured Output ---
class TitleSlide(BaseModel):
"""Title slide model"""
title: str = Field(description="Main title of the presentation")
subtitle: str = Field(description="Subtitle or description")
class TableData(BaseModel):
"""Table data model"""
headers: List[str] = Field(description="Table column headers")
rows: List[List[str]] = Field(description="Table rows data")
class BulletSlide(BaseModel):
"""Bullet point slide model"""
heading: str = Field(description="Slide heading")
bullet_points: List[str] = Field(description="List of bullet points, use >> prefix for step-by-step slides")
img_keywords: Optional[str] = Field(default="", description="Keywords for image search (not URLs)")
class TableSlide(BaseModel):
"""Table slide model"""
heading: str = Field(description="Slide heading")
table: TableData = Field(description="Table data with headers and rows")
img_keywords: Optional[str] = Field(default="", description="Keywords for image search (not URLs)")
class PresentationSlides(BaseModel):
"""Complete presentation model"""
title_slide: TitleSlide = Field(description="First slide must be a title slide")
content_slides: List[Union[BulletSlide, TableSlide]] = Field(
description="Content slides including bullet slides and table slides"
)
# Load camel logo as base64
def get_base64_img(path, w=32):
with open(path, "rb") as img_f:
b64 = base64.b64encode(img_f.read()).decode()
return f"<img src='data:image/png;base64,{b64}' width='{w}' style='vertical-align:middle;margin-bottom:4px;'>"
camel_logo = get_base64_img("assets/CAMEL_logo.jpg", 40)
# --- Page config ---
st.markdown(
f"""
<div style='display:flex; align-items:center; gap:12px;'>
<span style="font-size:2.2rem;font-weight:700;">PPTX Generator using pptx_toolkit with</span>
{camel_logo}
</div>
""",
unsafe_allow_html=True
)
# --- Sidebar Config ---
st.sidebar.header("āļø Configuration")
openai_key = st.sidebar.text_input(
"OpenAI API Key (for ChatAgent/OpenAI backend)", type="password",
help="Get yours at https://platform.openai.com/account/api-keys"
)
pexels_key = st.sidebar.text_input(
"Pexels API Key", type="password", help="(Optional) Needed if you want images fetched from Pexels"
)
# --- User Input ---
st.header("1. Enter Your Presentation Topic")
topic = st.text_input("Topic (e.g., āBlockchain for Beginnersā)")
slide_count = st.slider("Number of slides (excluding title slide)", 3, 10, 5)
# --- Construct the JSONāgeneration instructions ---
def pptx_prompt(topic: str, slide_count: int) -> str:
r"""Create a prompt for structured output generation"""
return f"""
Create a presentation about "{topic}" with exactly {slide_count + 1} slides total (1 title slide + {slide_count} content slides).
Requirements:
1. First slide must be a title slide with appropriate title and subtitle
2. Include at least one step-by-step slide (bullet points starting with ">>")
3. Include at least one table slide with relevant data
4. At least TWO content slides must have meaningful img_keywords for visual content
5. Use Markdown formatting (**bold**, *italic*) in bullet points
6. Make content clear, concise, and engaging for the topic "{topic}"
The presentation should cover key aspects of {topic}, including practical information, processes, and relevant data.
"""
# --- Generate Slide JSON via ChatAgent ---
def generate_pptx_json_with_agent(topic: str, slide_count: int, api_key: str):
# 1. Build the full instruction prompt
full_prompt = pptx_prompt(topic, slide_count)
# 2. Instantiate a ChatAgent pointing at OpenAI GPT-4o
os.environ["OPENAI_API_KEY"] = api_key
agent = ChatAgent(
system_message="You are an AI Assistant that strictly follows instructions to produce only valid JSON for slides.",
message_window_size=5,
model=ModelFactory.create(
model_platform=ModelPlatformType.OPENAI,
model_type=ModelType.GPT_4_1,
model_config_dict={"temperature": 0.0},
)
)
# 3. Call the agent with our instruction prompt
try:
response = agent.step(full_prompt,response_format=PresentationSlides)
# Parse JSON string into Pydantic object
json_content = response.msgs[0].content
presentation_data = PresentationSlides.model_validate_json(json_content)
# Convert to the expected JSON format
slides_json = []
# Add title slide
slides_json.append({
"title": presentation_data.title_slide.title,
"subtitle": presentation_data.title_slide.subtitle
})
# Add content slides
for slide in presentation_data.content_slides:
if isinstance(slide, BulletSlide):
slides_json.append({
"heading": slide.heading,
"bullet_points": slide.bullet_points,
"img_keywords": slide.img_keywords or ""
})
elif isinstance(slide, TableSlide):
slides_json.append({
"heading": slide.heading,
"table": {
"headers": slide.table.headers,
"rows": slide.table.rows
},
"img_keywords": slide.img_keywords or ""
})
return slides_json, None
except Exception as e:
return None, f"ā Structured output generation failed: {e}"
# --- Build & Return PPTX Bytes ---
def build_pptx(slides):
os.environ["PEXELS_API_KEY"] = pexels_key or ""
pptx_toolkit = PPTXToolkit(output_dir="outputs")
out_name = f"demo_pptx_{abs(hash(json.dumps(slides)))}.pptx"
result = pptx_toolkit.create_presentation(json.dumps(slides), out_name)
pptx_path = os.path.join("outputs", out_name)
if os.path.exists(pptx_path):
with open(pptx_path, "rb") as f:
return f.read(), out_name, result
else:
return None, None, result
# --- Main App Logic ---
if not openai_key:
st.info("Please enter your OpenAI API key in the sidebar to continue.")
st.stop()
if topic and st.button("Generate Presentation"):
st.info("š Generating slide JSON with ChatAgent (GPT-4.1)...")
slides, error = generate_pptx_json_with_agent(topic, slide_count, openai_key)
if error:
st.error(error)
st.stop()
st.success("ā
Slide JSON created! Now building PPTX...")
pptx_bytes, fname, pptx_result = build_pptx(slides)
if pptx_bytes:
st.success("š Your presentation is ready!")
st.download_button(
label=f"Download {fname}",
data=pptx_bytes,
file_name=fname,
mime="application/vnd.openxmlformats-officedocument.presentationml.presentation"
)
else:
st.error(f"PPTX generation failed: {pptx_result}")
st.markdown("---")
st.caption("Made with ā¤ļø using CAMEL-AI, OpenAI & PPTXToolkit")