How ashrokss/ai-agents works
Project type and purpose
• This is a Python monorepo of multiple AI agents built on the Agno framework.
• Each subfolder (Deep Research Agent, Finance Agent, Recruitment Agent Team, SQL-Agent) is a standalone “agent” project containing its own entrypoint (app.py or app-st.py), agent logic, utilities, and dependencies.
• The root README describes the overall goal: to provide a baseline for LLM-powered, tool-enabled multi-agent systems.
File structure
• Top-level folders:
• Deep Research Agent/
• Finance Agent/
• Recruitment Agent Team/
• SQL-Agent/
• Common patterns in each agent folder:
• README.md – describes that agent’s use case
• app.py or app-st.py – application entrypoint (API server or Streamlit UI)
• agents.py or finance_agent.py – agent orchestration code
• utils/ – shared helper functions (Recruitment Agent Team)
• config.py (SQL-Agent) – service configs, database credentials, etc.
• requirements.txt – Python dependencies
README for additional insights
• Agno provides the core abstractions: AgentManager, Tool integration, LLM client adapters (OpenAI, Azure, Gemini).
• It supports streaming, tool chaining, multi-agent workflows—key architectural capabilities.