adding pdf-tex summarizer example

agents/crypto_trader_bot/README.md

@@ -25,7 +25,7 @@ Inspired by [How to build your own Add Liquidity Signal Telegram Bot for Solana
 1. Clone the repository:
 ```bash
 git clone https://github.com/swarmzero/examples.git
-cd examples/agents/crypto-trader-bot
+cd examples/agents/crypto_trader_bot
 ```
 
 2. Install dependencies using Poetry:

workflows/pdf_summarizer_tex_workflow/.env.example

@@ -0,0 +1,6 @@
+MODEL=gpt-4o-mini
+ENVIRONMENT=dev
+OPENAI_API_KEY=
+OPENROUTER_MODEL=gryphe/mythomax-l2-13b
+SWARMZERO_LOG_LEVEL=INFO
+LANGTRACE_API_KEY=

workflows/pdf_summarizer_tex_workflow/.gitignore

@@ -0,0 +1,3 @@
+.venv
+swarmzero-data
+.env

workflows/pdf_summarizer_tex_workflow/README.md

@@ -0,0 +1,39 @@
+# PDF Summarizer to Tex Workflow
+
+A sequential workflow built using SwarmZero framework that enables concise summarization of large pdf's to be outputted in a tex file 
+
+## Description
+
+This workflow utilizes PyMuPDF for extracting text out of a pdf and is built on top fo the SwarmZero framework, providing enhanced summarization cpabilities wiht AI-powered processing. 
+
+## Prerequisites
+
+- Python 3.11 or higher
+- Poetry package manager
+- OpenAI API Key
+- Langtrace API Key
+
+## Installation
+
+1. Clone the repository:
+```bash
+git clone https://github.com/swarmzero/examples.git
+cd examples/workflows/pdf_summarizer_tex_workflow
+```
+
+2. Install dependencies using Poetry:
+```bash
+poetry install --no-root
+```
+
+3. Set up environment variables:
+Create a `.env` file in the root directory and add your API keys based on the .env.example file
+## Usage
+
+Run the workflow on a pdf like so:
+``` bash
+python main.py path/to/pdf/text.pdf
+```
+
+## Learn more
+Visit [SwarmZero](https://swarmzero.ai) to learn more about the SwarmZero framework.

workflows/pdf_summarizer_tex_workflow/agents/__init__.py

@@ -0,0 +1,4 @@
+from .formatter_agent import formatter_agent
+from .summarizer_agent import summarizer_agent
+
+__all__ = ["formatter_agent", "summarizer_agent"]

workflows/pdf_summarizer_tex_workflow/agents/formatter_agent.py

@@ -0,0 +1,20 @@
+from swarmzero.agent import Agent, SDKContext
+import os
+from tools import format_latex
+
+CONFIG_PATH = os.path.abspath(
+    os.path.join(
+        os.path.dirname(__file__), 
+        os.pardir,                   
+        "swarmzero_config.toml"    
+    )
+)
+sdk_context = SDKContext(CONFIG_PATH)
+formatter_agent = Agent(
+    name="formatter",
+    functions=[format_latex],
+    instruction="Format bullet points into LaTeX.",
+    description="Wraps bullets into a LaTeX itemize document.",
+    sdk_context=sdk_context,
+    chat_only_mode=True,
+)

workflows/pdf_summarizer_tex_workflow/agents/summarizer_agent.py

@@ -0,0 +1,20 @@
+from swarmzero.agent import Agent, SDKContext
+import os
+from tools import summarize_bullets
+
+CONFIG_PATH = os.path.abspath(
+    os.path.join(
+        os.path.dirname(__file__), 
+        os.pardir,                   
+        "swarmzero_config.toml"    
+    )
+)
+sdk_context = SDKContext(CONFIG_PATH)
+summarizer_agent = Agent(
+    name="summarizer",
+    functions=[summarize_bullets],
+    instruction="Summarize text into bullet points.",
+    description="Returns bullet-point summary for a given text chunk.",
+    sdk_context=sdk_context,
+    chat_only_mode=True,
+)

workflows/pdf_summarizer_tex_workflow/main.py

@@ -0,0 +1,66 @@
+import asyncio
+import os
+import sys
+from typing import List
+
+from swarmzero.sdk_context import SDKContext
+from swarmzero.workflow import Workflow, WorkflowStep, StepMode
+
+from agents import formatter_agent, summarizer_agent
+from util import extract_text_from_pdf, chunk_text
+
+CONFIG_PATH = os.path.join(os.path.dirname(__file__), "swarmzero_config.toml")
+sdk_context = SDKContext(CONFIG_PATH)
+
+async def run_summarizer(chunk: str, **kwargs) -> str:
+    return await summarizer_agent.chat(chunk)
+
+async def run_formatter(bullets: str, **kwargs) -> str:
+    return await formatter_agent.chat(bullets)
+
+workflow = Workflow(
+    name="pdf_latex_summary",
+    instruction="Summarize PDF and convert to LaTeX.",
+    description="Runs summarizer then formatter agents sequentially.",
+    steps=[
+        WorkflowStep(name="SummarizeBullets",
+                     runner=run_summarizer,
+                     mode=StepMode.SEQUENTIAL),
+
+        WorkflowStep(name="FormatLatex",
+                     runner=run_formatter,  
+                     mode=StepMode.SEQUENTIAL),
+    ],
+    sdk_context=sdk_context,
+)
+
+async def main():
+    pdf_path = sys.argv[1] if len(sys.argv) > 1 else "sample.pdf"
+    text = extract_text_from_pdf(pdf_path)
+    chunks = chunk_text(text)
+
+    all_items: List[str] = []
+    for idx, chunk in enumerate(chunks, start=1):
+        print(f"Chunk {idx}/{len(chunks)} → processing…")
+        latex_doc = await workflow.run(chunk)
+        all_items += [line.strip() for line in latex_doc.splitlines() if line.strip().startswith("\\item")]
+
+    final = (
+        "\\documentclass{article}\n"
+        "\\usepackage[utf8]{inputenc}\n"
+        "\\usepackage{enumitem}\n"
+        "\\begin{document}\n"
+        "\\section*{Summary Notes}\n"
+        "\\begin{itemize}[leftmargin=*, label=--]\n"
+        + "\n".join(all_items) +
+        "\n\\end{itemize}\n"
+        "\\end{document}"
+    )
+
+    out_file = os.path.splitext(os.path.basename(pdf_path))[0] + "_summary.tex"
+    with open(out_file, "w", encoding="utf-8") as f:
+        f.write(final)
+    print(f"Written LaTeX summary to {out_file}")
+
+if __name__ == "__main__":
+    asyncio.run(main())