Skip to content

DarshanS-Dev/Kodo

Repository files navigation

Kōdo — The AI Coding Mentor That Remembers You

Most AI tutors forget you the moment you close the tab. Kōdo doesn't. Kōdo is an AI-powered coding mentor with persistent memory — it remembers your learning patterns, your weak spots, and the explanations that actually clicked for you, then uses that to become a fundamentally different tutor the longer you use it.

Live Demo

  • Website: [coming soon]
  • Demo Video: [coming soon]

Features

  • Comeback Brief — Every session opens with a personalized summary of where you left off
  • Adaptive Explanations — Kōdo remembers which analogies worked for you and uses them again
  • Pattern Interrupts — Kōdo detects recurring behavioral patterns and calls them out before you repeat them
  • Built-in Code Editor — Monaco Editor (same as VS Code) with syntax highlighting and code execution via Piston API
  • Problem Recommendation — Kōdo suggests problems based on your weak areas and tracks what you've already solved
  • Weekly Insight Card — Behavioral analysis of your learning patterns, not just topics covered

How Hindsight is Used

Kōdo uses all three Hindsight API calls meaningfully:

  • retain() — Called after every chat message and problem attempt. Stores behavioral inferences — not just what was said, but what it reveals about how the user thinks and learns.
  • recall() — Called before every response to fetch relevant memory context. When a user is solving a problem, recall fetches their history with that specific problem type.
  • reflect() — Called for the Comeback Brief, Pattern Interrupts, problem recommendations, and the Weekly Insight Card. This is where Hindsight reasons over the full observation history to generate non-obvious insights about the user.

Tech Stack

Layer Technology
Frontend Next.js + Tailwind CSS
Code Editor Monaco Editor
Code Execution Piston API
Backend FastAPI
Memory Hindsight Cloud
LLM Groq (llama-3.3-70b-versatile)

Setup — Run Locally

Backend

cd Backend
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Create a .env file in the Backend/ folder:

HINDSIGHT_API_KEY=your_key_here
HINDSIGHT_BASE_URL=your_instance_url_here
GROQ_API_KEY=your_key_here

Run the server:

uvicorn main:app --reload

Frontend

npm install
npm run dev

Update BASE_URL in src/app/workspace/page.tsx to point to your backend URL.

API Endpoints

Method Endpoint Description
POST /session/start Generate Comeback Brief for returning user
POST /chat Main conversation loop with memory
POST /problem/submit Submit code attempt, retain behavioral analysis
GET /problem/list Fetch all problems for sidebar
GET /problem/{id} Fetch single problem
GET /insight/weekly Generate weekly behavioral insight

Team

  • Darshan — Backend, Hindsight integration
  • Arjun — Frontend, UI/UX

Built With

About

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors