🌟 AI Services Agency — Your Tiny Team of Super-Helpful AI Buddies! 🤝🚀
Make a smart plan for your app or startup idea with a friendly team of AI helpers. Each AI plays a role (like CEO, CTO, Product, Developer, and Client Success) and works together to give you a full plan — strategy, tech, roadmap, and go-to-market — all inside a simple Streamlit app.
— Super simple. Super guided. Super powerful. 💥
✨ Quick Buttons
- ▶️ Run Locally:
streamlit run agency.py - 🔑 Get API Key: https://platform.openai.com/api-keys
- 📦 Install Requirements:
pip install -r requirements.txt
🎬 What This App Does (Like You’re 5!)
- You tell us your idea (what you want to build, how fast, and your budget).
- Five AI friends talk to each other to help you:
- CEO gives a big-picture plan 👨💼
- CTO picks the tech and architecture 🧠
- Product Manager plans features and roadmap 📍
- Developer explains how to build it 🛠️
- Client Success shows how to get customers 📣
- You get a neat, organized plan in tabs you can read and use.
It’s like a smart mini-agency in your laptop! 💻✨
🧩 What’s Inside
agency.py— The Streamlit app and all the agents/tools.requirements.txt— What to install.
No extra setup files. No hidden tricks. Just run it.
🚀 Setup — Step by Step (Zero Stress)
-
Install Python (version 3.10 or newer)
- Download: https://www.python.org/downloads/
- During install, check “Add Python to PATH”.
-
Get your OpenAI API Key
- Go to: https://platform.openai.com/api-keys
- Click “Create new secret key”.
- Copy it. Keep it private. You’ll paste it in the app.
-
Open a terminal in the project folder
- Example on Windows: Right-click → “Open in Terminal”.
-
(Optional but recommended) Create a virtual environment
bash
python -m venv .venv .venv\Scripts\activate
5) **Install the packages** ```bash pip install -r requirements.txt
- Run the app
bash
streamlit run agency.py
7) **Paste your OpenAI API key in the left sidebar** - Look for the box that says “OpenAI API Key”. - Paste your key. You’ll see “API Key accepted!”. - If you don’t paste a key, the app will ask you to add one and pause. 8) **Fill the form and click “Analyze Project”** - Enter name, description, project type, budget, priority, etc. - The agents will think, plan, and show results in neatly labeled tabs. That’s it. You did it! 🎉 --- ## 🧠 Features (What the AIs Do) - **CEO (Project Director)** - Runs the `AnalyzeProjectRequirements` tool first. - Evaluates complexity, timeline, feasibility, and key requirements. - Shares a strategic view of your project. - **CTO (Technical Architect)** - Waits for the CEO’s analysis. - Runs the `CreateTechnicalSpecification` tool. - Chooses architecture (monolithic/microservices/serverless/hybrid), tech stack, scalability. - **Product Manager** - Suggests features and a delivery roadmap. - **Lead Developer** - Explains how to build it, estimates, stack choices, and feasibility. - **Client Success Manager** - Provides go-to-market and customer acquisition strategy. - **Async teamwork** - Uses `Agency(async_mode='threading')` for efficient multi-agent collaboration. --- ## 🏗️ How It Works (Under the Hood) - Built with `streamlit` for the UI. - Uses `agency-swarm` to create agents that can use tools and talk via an agency. - Two custom tools power structured thinking: - `AnalyzeProjectRequirements` - `CreateTechnicalSpecification` - Results are displayed in five tabs: - CEO’s Project Analysis - CTO’s Technical Specification - Product Manager’s Plan - Developer’s Implementation - Client Success Strategy --- ## 📋 Form Fields You’ll See - Project Name - Project Description - Project Type (Web App, Mobile App, API, Data, AI/ML, Other) - Timeline (1–2, 3–4, 5–6, 6+ months) - Budget ($10k–$25k, $25k–$50k, $50k–$100k, $100k+) - Priority (High / Medium / Low) - Optional: Technical requirements and special considerations --- ## 🔐 Privacy & Keys - Your API key is entered locally in the Streamlit sidebar. - The app calls `set_openai_key(key)` and uses it to authorize agent calls. - Don’t share your key. Treat it like a password. --- ## 🧪 Test It Fast Try this minimal input: - Project Name: SuperTodo - Description: A simple web app to manage tasks with sharing. - Type: Web Application - Timeline: 3–4 months - Budget: $25k–$50k - Priority: High Click “Analyze Project” and read each tab’s output. --- ## 🛠️ Troubleshooting - **App says: Please enter your OpenAI API Key** - Open the left sidebar and paste your key. Use a valid, active key. - **Python/Package errors on install** - Ensure Python 3.10+. - Upgrade pip: `python -m pip install --upgrade pip` - Reinstall: `pip install -r requirements.txt` - **Streamlit not found** - Install requirements again or run: `pip install streamlit` - **Agency Swarm errors** - Installed version is pinned in `requirements.txt` (`agency-swarm==0.4.1`). - Reinstall just in case: `pip install --force-reinstall agency-swarm==0.4.1` - **Nothing shows or the app reloads** - Check the terminal for errors. - Try restarting the app: stop with Ctrl+C, then `streamlit run agency.py` again. --- ## 📦 Dependencies From `requirements.txt`: - python-dotenv==1.0.1 - agency-swarm==0.4.1 - streamlit --- ## 📚 FAQ - **Do I need Docker?** - No. Just Python and pip. - **Where do I put the API key?** - In the app’s left sidebar. No `.env` file is needed. - **Can I change the agents?** - Yes, open `agency.py` and adjust agent names, instructions, or tools. - **Can I use Claude, Gemini, etc.?** - The app is wired for OpenAI via `agency-swarm`. Other models would require code changes. - **Where does my data go?** - Inputs are processed locally, and requests go to OpenAI with your key. --- ## 🧭 Roadmap Ideas - Save/export analyses as Markdown or PDF - Add example presets for common project types - Optional `.env` support for API keys - Theming and dark mode polish --- ## 🙌 Credits - Built with ❤️ using [Streamlit](https://streamlit.io) and [Agency Swarm](https://pypi.org/project/agency-swarm/). --- ## 🔗 Demo (Video) If you prefer a quick look, here’s a demo link: https://github.com/user-attachments/assets/a0befa3a-f4c3-400d-9790-4b9e37254405 --- ## 🧾 License No license file is included in this repository at the moment.