How tarunteja-nagandla/ai-task-manager works
This is a Streamlit natural-language task manager. A user submits a chat request; the UI passes it to a LangChain/LangGraph agent, which adds date context and can dispatch task CRUD tools, then returns a response to the chat. The tools persist tasks through SQLAlchemy in SQLite. The README also describes Groq-backed responses, in-memory agent checkpointing, task filtering, proactive completion detection, and chat-history controls. Source excerpts show the main calls and persistence path; the README documents the agent capabilities and integrations.
Main entry points: agent.py Agent runtime · database.py Task ORM · app.py Chat UI · tools.py Create task
Components 12
Chat interface
- Chat UI
app.py
Agent orchestration
- Agent runtime
agent.py - Date context
agent.py - Agent memory
agent.py
Task operations
- Create task
tools.py - List and filter
tools.py - Update task
tools.py - Delete task
tools.py
Persistence
- Task ORM
database.py - SQLite tasks
database.py
Other components
- User
- Groq LLM
Connections 16
- User → Chat UI: submits request
README.md - Chat UI → Agent runtime: calls agent
app.py - Chat UI → Task ORM: initializes schema
app.py - Agent runtime → Date context: adds date
agent.py - Agent runtime → Agent memory: configures checkpoint
agent.py - Agent runtime → Groq LLM: queries model
README.md - Agent runtime → Create task: dispatches tool
agent.py - Agent runtime → List and filter: dispatches tool
agent.py - Agent runtime → Update task: dispatches tool
agent.py - Agent runtime → Delete task: dispatches tool
agent.py - Agent runtime → Task ORM: initializes schema
agent.py - Create task → Task ORM: creates task
tools.py - List and filter → Task ORM: queries tasks
tools.py - Update task → Task ORM: updates task
tools.py - Delete task → Task ORM: deletes task
tools.py - Task ORM → SQLite tasks: persists records
database.py