AI agent workflows designed to save you time when developing products - across research, development, testing, marketing, and sales.
30 Agents in 30 Days is a challenge to create and share 30 practical AI agent workflows that help product builders save time and improve their process. Each day, a new agent workflow is released to help with different aspects of the product development lifecycle.
Each agent workflow is contained in its own folder with complete documentation, code, and instructions for implementation.
/
├── news/ # Day 01: News summarization (PocketFlow)
├── 02-human-alignment/ # Day 02: Human alignment (SmolAgents)
├── 03-ai-vs-ai-benchmark/ # Day 03: AI vs AI Benchmark (optional)
├── 04-trends/ # Day 04: Trend detection (Gradio)
├── 05-whatsapp/ # Day 05: WhatsApp multi-agent (SmolAgents)
├── 06-knowledge-graph/ # Day 06: Knowledge graph explorer (SmolAgents)
├── 07-project-management/ # Day 07: Project management with Notion (SmolAgents)
├── 08-whiteboarding/ # Day 08: Whiteboarding with FigJam (SmolAgents)
├── 09-google-sheets/ # Day 09: Google Sheets integration (SmolAgents)
├── 10-documentation-gen/ # Day 10: Documentation generation (SmolAgents)
├── 11-ui/ # Day 11: No-code multi-agent builder (Streamlit)
├── 12-whatsapp-api/ # Day 12: WhatsApp Business API integration (advanced)
├── mcp-servers/ # Backend MCP connectors (WhatsApp, Notion, Google Sheets, etc.)
└── assets/ # Shared assets
Every day for 30 days, I'm releasing a new AI agent workflow to help you build better products faster. The agents are organized into four main categories:
Agents that help you gather insights and analyze data, including:
- Market research automation
- Competitive analysis
- User feedback processing
- Trend identification
Agents that accelerate your coding and development workflow, such as:
- Code generation and refactoring
- Architecture planning
- Documentation creation
- Feature prioritization
Agents that help you test your products thoroughly, including:
- Automated QA workflows
- User journey simulation
- Performance testing
- Security assessment
Agents that help you promote and sell your products, such as:
- Content generation
- Ad copy optimization
- Lead qualification
- Sales outreach automation
Each agent workflow folder includes a README.md tutorial with step-by-step instructions and examples.
- Browse: Explore the different agent workflows organized by category
- Implement: Follow the instructions in each agent's README to set up the workflow
- Customize: Adapt the agents to your specific product needs
- Contribute: Share your improvements or ideas by opening issues or PRs
To help you dive in, here’s a suggested order for exploring the workflows:
- Day 11: No-Code Multi-Agent Builder (
11-ui) Try the live demo at https://mechifact.streamlit.app/ or run locally:pip install -r requirements.txt streamlit streamlit run 11-ui/main.py
- Day 01: News Summarization with PocketFlow (
news)python news/main.py
- Day 02: Human Alignment (
02-human-alignment) Uses SmolAgents and the arXiv API:python 02-human-alignment/main.py
- Day 03: AI vs AI Benchmark (
03-ai-vs-ai-benchmark) Optional A conceptual hackathon project—explore at your leisure. - Day 04: Trend Detection (Gradio) (
04-trends)pip install -r 04-trends/requirements.txt python 04-trends/interface.py
- Day 05: WhatsApp Multi-Agent (
05-whatsapp) Requires setup (see folder README):python 05-whatsapp/main.py
- Day 06: Knowledge Graph Explorer (
06-knowledge-graph)python 06-knowledge-graph/interface.py
- Day 07: Project Management with Notion (
07-project-management)python 07-project-management/main.py
- Day 08: Whiteboarding with FigJam (
08-whiteboarding) Requires FigJam plugin and local server (see folder README):python 08-whiteboarding/main.py
- Day 09: Google Sheets Integration (
09-google-sheets)python 09-google-sheets/main.py
- Day 10: Documentation Generation (
10-documentation-gen) Compare single vs multi-agent:python 10-documentation-gen/single_agent.py python 10-documentation-gen/multi_agent.py
- Day 12: WhatsApp Business API Integration (
12-whatsapp-api) Advanced Requires a WhatsApp Business account and setup. See 12-whatsapp-api/README.md for details.
To ensure you also clone all MCP server backends (Notion, Google Sheets, WhatsApp), use the --recursive flag when cloning:
git clone --recursive https://github.com/yourusername/agents.gitIf you've already cloned without --recursive, initialize and fetch the submodules with:
git submodule update --init --recursiveEach workflow includes entry points (e.g., main.py, interface.py). From the repository root, install dependencies and run any workflow:
# Install root dependencies
pip install -r requirements.txt
# (For Trend detection)
pip install -r 04-trends/requirements.txt
# (For Streamlit UI)
pip install streamlit
# Day 01: News Summarization (PocketFlow)
python news/main.py
# Day 02: Human Alignment (SmolAgents)
python 02-human-alignment/main.py
# Day 03: AI vs AI Benchmark (optional)
python 03-ai-vs-ai-benchmark/main.py
# Day 04: Trend Detection (Gradio)
python 04-trends/interface.py
# Day 05: WhatsApp Multi-Agent (SmolAgents)
python 05-whatsapp/main.py
# Day 06: Knowledge Graph Explorer (SmolAgents)
python 06-knowledge-graph/interface.py
# Day 07: Project Management with Notion (SmolAgents)
python 07-project-management/main.py
# Day 08: Whiteboarding with FigJam (SmolAgents)
python 08-whiteboarding/main.py
# Day 09: Google Sheets Integration (SmolAgents)
python 09-google-sheets/main.py
# Day 10: Documentation Generation (SmolAgents)
python 10-documentation-gen/single_agent.py
python 10-documentation-gen/multi_agent.py
# Day 11: No-Code Multi-Agent Builder (Streamlit)
streamlit run 11-ui/main.py
# Day 12: WhatsApp Business API Integration (advanced)
# See 12-whatsapp-api/README.md for setup instructions- Follow #30daysofagents @leeps.bsky.social on Bluesky for daily updates
- Visit agents.mechifact.com for interactive demos