How wangzihe0326/matfinder-ai works
MatFinder AI takes a user’s material requirements through local catalog search and evidence-aware recommendation, then supports material details, comparisons, optional AI explanations, and report export. The browser app is the user-facing orchestrator; the Node server owns API requests, repository reads, and optional OpenAI calls. SQLite is the documented source of truth. README capabilities inform the page, comparison, and export nodes; sampled excerpts do not establish every UI-to-module call or all evidence-validation wiring, so uncertain links are omitted.
Main entry points: app.js Multi-page app · server.js HTTP API server · material-repository.js Material repository · evidence-model.js Evidence model
Components 19
Browser experience
- Multi-page app
public/app.js - Material catalog
public/catalog-layer.js - Polymer families
public/polymer-families.js - Material detail
public/app.js - Compare workflow
public/app.js - AI explanations
public/app.js - Selection report
public/app.js
Catalog and matching
- Catalog search
public/catalog-search.js - Recommendation engine
public/recommendation-engine.js - Curated material source
data/materials.js
API and services
- HTTP API server
server.js - Material repository
material-repository.js - SQLite catalog
- API protection
api-protection.js
Evidence and quality
- Evidence model
evidence-model.js - Evidence validation
evidence-validator.js - Material quality
material-quality.js
Other components
- Material selector
- OpenAI API
Connections 21
- Material selector → Multi-page app: enters requirements
- Multi-page app → Catalog search: searches catalog
- Multi-page app → Recommendation engine: ranks requirements
- Multi-page app → Material catalog: browses catalog
- Multi-page app → Polymer families: searches families
- Multi-page app → HTTP API server: requests materials
- HTTP API server → Material repository: queries records
- Material repository → SQLite catalog: reads catalog
- Curated material source → SQLite catalog: source of truth
- Multi-page app → Material detail: opens selection
- Multi-page app → Compare workflow: selects materials
- Multi-page app → AI explanations: requests explanation
- AI explanations → HTTP API server: posts analysis
- HTTP API server → Material repository: loads profiles
- HTTP API server → OpenAI API: requests narrative
- Multi-page app → HTTP API server: posts comparison
- HTTP API server → OpenAI API: requests comparison
- HTTP API server → API protection: checks access
- Multi-page app → Selection report: exports report
- Material repository → Evidence model: normalizes evidence
- Material quality → Evidence model: uses definitions