Skip to content

codepath/pathreview-test

 
 

Repository files navigation

PathReview

AI-powered portfolio review assistant that helps early-career developers strengthen their professional portfolios.

PathReview analyzes GitHub profiles, resumes, and project repositories to generate structured, actionable feedback on portfolio completeness, project quality, skill gaps, and presentation improvements.

Features

  • Profile Ingestion — Upload a resume (PDF or Markdown), connect a GitHub profile, and link project repositories
  • RAG-Powered Feedback — Retrieval-augmented generation produces specific, evidence-based feedback referencing your actual work
  • Multi-Tool Agent — An AI agent orchestrates GitHub analysis, skill extraction, README scoring, and market comparison
  • Safety Guardrails — Bias detection, content filtering, PII scrubbing, and prompt injection defense ensure feedback is constructive and safe
  • Web Dashboard — View results, track improvement over time, and export shareable review summaries

Quick Start

Windows users: Use Git Bash to run these commands, not PowerShell. See docs/SETUP.md for Windows-specific setup including installing make.

# Clone and enter the repo
git clone https://github.com/jamjamgobambam/pathreview.git
cd pathreview

# Configure environment (add your OPENROUTER_API_KEY to .env)
cp .env.example .env

# Start backing services — must be running before make setup
docker compose up -d

# Run first-time setup (installs deps, runs migrations, seeds DB, installs frontend)
make setup

# Start the application
make run

Then open http://localhost:5173 in your browser.

For detailed setup instructions including platform-specific notes, see docs/SETUP.md.

Architecture

PathReview is structured as a multi-service Python + React application with five major subsystems:

Subsystem Directory Description
API Layer api/ FastAPI REST API with authentication, validation, and rate limiting
Ingestion Pipeline ingestion/ Document parsing, chunking, and embedding generation
RAG System rag/ Hybrid retrieval, LLM-based review generation, and quality evaluation
Agent System agent/ Multi-tool orchestration with planning, state management, and error handling
Safety Layer safety/ Content filtering, bias detection, PII scrubbing, and prompt defense
Frontend frontend/ React + TypeScript dashboard with Vite

For a detailed architecture overview, see docs/ARCHITECTURE.md.

Contributing

We welcome contributions! Please read docs/CONTRIBUTING.md before submitting a pull request.

Development

make help          # Show all available commands
make test-unit     # Run unit tests (~30 seconds)
make check         # Run linter + formatter + type checker
make run           # Start the dev servers

License

MIT

About

AI-powered portfolio review assistant — AI 201 Module 3 simulated contribution project

Resources

Contributing

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 86.5%
  • TypeScript 12.6%
  • Other 0.9%