AI Agents
This project is a collection of AI agents built for learning and experimenting with AI agent development. Each agent is designed to explore different aspects of autonomous systems, from simple task automation to complex reasoning and decision-making. The idea is to prove out different capabilities that can be used in combination for more complex projects.
Note: Most of these do not fit the more common definition of AI Agent, in that they do not do all actions autonomously. That part will come later with more advanced agent design and implementation.
Available AI Agents
Daily Standup Agent
Automates daily standup meeting preparation and summaries using natural language processing and structured outputs.
Top News Agent
Fetches and summarizes the top news stories of the day using web search capabilities, summarization, and markdown table output.
Bug Report Agent
Interactive bug reporting agent that creates structured bug reports through multi-turn conversation, conversation state management, and input validation with Pydantic model outputs.
Dev Tools Agent
OpenAI-powered developer automation agent featuring natural language chat and custom function calling with multiple functions.
File Search Agent
Ask questions about one or more text files using AI-powered analysis.
CLI Interface
The project includes a command-line interface that provides access to all agents:
- news: Fetch top news stories (
./ai news 5
) - standup: Format daily standup updates (
./ai standup "status message"
) - bug-report: Interactive bug reporting session (
./ai bug-report
) - file-search: Query text files (
./ai file-search FILE1 "question"
) - dev-tools: Interactive developer tools chat (
./ai dev-tools
)
Tech Stack
- Python - Primary language for all agent implementations
- OpenAI API - Core AI capabilities and language models
- Click - CLI framework for command-line interface
- Pydantic - Data validation and structured outputs
Key Learning Areas
- Agent orchestration and workflow management
- Multi-turn conversation handling
- Structured data output generation
- Tool calling and function integration
- Interactive CLI development
- AI-powered automation patterns
Future Enhancements
- Support for non-plain-text file types (PDF, DOCX, CSV) in File Search Agent
- CLI-level integration tests for all commands
- Additional agent types and feature enhancements